The conventional techniques for preparing fiber gratings involve ultraviolet exposure and CO2 laser methods. The ultraviolet exposure is advantageous due to its simplicity and ease of alignment. However, it typically requires hydrogen-carrying sensitization treatment on the fiber and the refractive index modulation can be easily erased at high temperatures, limiting its applicability in extreme environments. On the other hand, the CO2 laser method is commonly used for producing long-period fiber gratings but its measurement sensitivity is susceptible to high temperatures. To address the limitations of these traditional methods, femtosecond laser scribing technology has emerged. This technology encompasses femtosecond laser direct writing, femtosecond laser holographic interference and femtosecond laser phase mask methods. Among these, femtosecond laser direct writing offers high efficiency, low pulse energy requirements, and the ability to modulate the center wavelength, grating spacingand grating length based on sensing requirements. Moreover, the experimental setup for femtosecond laser direct writing is simple and does not require a phase mask to control the grating period. Femtosecond laser direct writing fiber gratings exhibit flexible refractive index modulation, excellent high-temperature performance, and high mechanical strength. As a result, they have found widespread applications in sensors, lasers, and other optical devices. This paper provides a brief introduction to the working principle and typical writing methods of femtosecond laser direct writing fiber gratings. It also summarizes the research progress of three direct writing methods of femtosecond laser, both domestically and internationally. The advantages and disadvantages of these three methods are compared and analyzed in terms of preparation efficiency and spectral quality. Additionally, the paper delves into the detailed analysis and discussion of spectral optimization methods, including laser pulse energy, grating length, fiber type, beam shaping, and grating apodization. The ultimate goal is to achieve a high reflection peak, narrow 3dB bandwidth, and low insertion loss.

Surface-enhance Raman scatting (SERS) has been widely used in biomedical detection due to its high sensitivity, non-invasive, multi-channel detection and other characteristics. The abnormal expression of microribonucleic acid （miRNAs） has been found to be associated with a variety of diseases, and miRNAs have become a novel biomarker. The development of simple, sensitive and reliable miRNAs detection methods is of great significance for studying biological function, medical diagnosis, disease treatment and targeted drug research of miRNAs. Nano-SERS probes combined with nucleic acid signal amplification strategy showed high sensitivity for miRNAs detection. However, the components in the actual samples are complex, resulting in background signal interference detection results, and the tedious separation and purification process increases the detection time. Researchers have recently combined other techniques to optimize detection, improve sample throughput, simplify operations, reduce analysis time, and improve resolution. This paper mainly introduces the latest progress in miRNAs detection method based on SERS technology, discusses the advantages and necessity of technology fusion, and summarises the existing problems of miRNAs detection based on SERS technology for clinical detection, aiming to provide a reference for the design of a new fast, sensitive and reliable miRNAs detection platform.

With the emergence of new food processing technology and materials, food contamination problems are also more frequent outbreaks of international food safety incidents. Foodborne diseases caused by food contamination seriously threaten peoples lives and health, and how to control food quality has become the focus of peoples attention. Terahertz technology isan emerging detection technologythatreflects the physical structure, properties, and chemical composition of the sample at the same time, access to the sample in the terahertz band of time and frequency domain information as a material “fingerprint” spectrum, used in the field of nondestructive testing and identification and has achieved certain results. Compared with the traditional means of detection, the advantages of terahertz technology, such as high safety, good visualization, and strong spectral resolution, give it great potential and application prospects in food contamination detection. This paper briefly describes the basic principles of terahertz spectroscopy, takes the most commonly used terahertz time-domain spectroscopy as the main research focus, summarizes the application of this technology in the detection of biological food contamination, chemical food contamination, and physical food contamination, and finally focuses on the current status of the practical application of terahertz technology in the field of food contamination detection, discusses the development trend and application prospects, and analyzes the current research still exists in the field of food contamination detection. Finally, it focuses on the practical application of terahertz technology in the field of food contamination detection, discusses its development trend and application prospects, and analyzes the problems that still exist in the current research to providereferences for the further development of terahertz technology in the field of food safety detection.

Glyoxal and methylglyoxal are two typical types of α-dicarbonyl compounds in the atmospheric environment. The concentration variation is an important characterization of the oxidation process and reaction activity of atmospheric VOCs, which is of great significance for studying the oxidation reaction of the atmosphere. However, the characteristics of extremely low concentrations, short lifetime, and strong activity of glyoxal and methylglyoxal bring out certain challenges to accurately detecting their concentrations, resulting in limited monitoring results of the field environment and a lack of research on the mechanism of atmospheric chemical reactions. Several methods have been developed for detecting α-dicarbonyl compounds, such as chemical derivatization and mass spectrometry, which can effectively achieve gas concentration monitoring, but the technology also has certain limitations. In recent years, with the advancement of optical technology, a series of spectral methods has developed, such as Differential Optical Absorption Spectroscopy, Cavity Enhanced Absorption Spectroscopy, Laser-Induced Phosphorescence, Fourier transform infrared, and other technologies, which have the characteristics of non-contact detection, low detection limit, high sensitivity, and high time resolution. This article summarizes the current status and development trends of spectral technology and provides a detailed explanation of the principle, key procedures, advantages, and disadvantages of the method. The article also lists the key features of the technology, such as device parameters, retrieval algorithm, detection limit, and relevant applications. At the same time, for the calibration requirements for the high activity α-dicarbonyl compounds, commonly used calibration methods, such as airflow dilution, temperature-controlled bubbler method, heating method, and atmospheric reaction method, were detailed and described and made a comparison. Finally, the field observation experiments of α-dicarbonyl compounds were summarized, including the experimental conditions, concentration results, and main conclusions, which indicated that spectral technology is a powerful tool for glyoxal and methylglyoxal detection. Some analysis of the correlation between concentration changes and primary pollution emissions, VOC oxidation, and secondary organic aerosol generation are carried out. The mixing ratio of formaldehyde, glyoxal, and methylglyoxal was mainly discussed, and the range of mixing ratio values under different environments was obtained. Some results indicate that RGF has lower values in BVOC environments, and a high mixing ratio may indicate the impact of artificial VOC sources.

Surface plasmon effects can be induced in nanoparticles of noble metals upon excitation of incident light, in which the light is coupled into free electrons, causing a collective oscillation of free electrons on the metal surface. When the oscillation frequency of free electrons is the same as that of the incident light, surface plasmon resonance may occur. The surface plasmon resonance effect of metal nanoparticles has various applications in many fields, including physics, chemistry, biology, etc. This work proposes a silver nanoparticle-based 1D composite array, and its surface plasmonic properties are investigated via the finite-difference time-domain method. In this structure, SiO2 is used as the substrate, and two silver nanoparticles are placed on the surface of the SiO2 substrate along the y-axis direction to form a 1D composite array with plane wave incident incident incident vertically along the negative z-axis direction. The results show that the surface plasmons of the 1D composite array are effectively excited in the wavelength range of 300~1 200 nm and that two plasmonic resonance peaks are revealed in the light absorption curve. The plasmonic resonances and their associated electromagnetic fields can systematically be tuned by adjusting the structural parameters of the array, including the nanoparticle size, the shape, and the period of the array. In addition, when one silver nanoparticle is fixed, and the size of the other particle is changed, it can be found that the two plasmonic resonance peaks of its absorption curve change differently. Combining the electric field diagrams, the equipartition excitonic resonance correlates to two different electromagnetic modes. By changing the shape of one of the silver nanoparticles, such as a pyramid, sphere, cylinder,orcube, it can be found that when the shape of the nanoparticle changes, the corresponding wavelength of the two plasmonic resonance peaks does not change, but the electromagnetic field distribution generated by different particle shapes is different, and the relative electromagnetic field intensity at the bottom is also significantly different. It can also be verified that the absorption curve has two different electromagnetic field modes. The results indicated in this work are significant in future research on designs of noble metal nanoparticle array-based plasmonic devices.

With the rapid development of light-emitting diodes, LED lighting technology is increasingly being used in various industries, and similar medical, printing, textile, beauty, and other industries require high color performance capability of LED lamps. Still, at the same time, the photobiological hazards brought about by LED lamps are also increasingly being paid attention to, among which is the LED blue light hazard, which is the most well-known. In recent years, many researchers have committed to reducing the blue light hazard of white LED. Still, most of them are in the case of low correlation color temperature to obtain the corresponding white light spectrum or through a section of the correlation color temperature range of the lowest blue light radiation efficiency of the correlation color temperature of the discrete exhaustive to find out the lowest blue light hazard of the spectra, can be obtained in different target correlation color temperature at the same time have a high color performance ability and low blue light radiation efficiency. The ability to obtain high color performance and low blue light hazard radiation efficiency at different target-relevant color temperatures of hybrid white LED spectra is almost non-existent. Based on the existing multiple LEDs, this paper proposes a nonlinear constrained optimization method with a simple objective function. Previous studies found that the blue light hazard-weighted radiant flux shows a linear relationship with its associated color temperature. As its associated color temperature increases, the blue light-weighted radiant flux also increases. Fix its relevant color temperature to obtain a white LED spectrum with low blue light hazard. You will get the blue light-weighted radiation flux corresponding to each targets relevant color temperature. According to the formula for evaluating the radiation efficiency of the blue light hazard of the LED light source, the numerator is fixed and only needs to maximize the denominator; the blue light hazard radiation efficiency will be minimized, so maximizing the radiation efficiency can be. For spectral power distribution, the spectral power distribution of the selected multiple LEDs is limited by using a filter, and at the same time, because of the actual production requirements to add a smoothing constraint on the spectral transmittance of the filter, and the smoothing constraint is accomplished by the method of weighted average. Finally, the radiation weight of each color LED is obtained by maximizing the “RP” method. The simulation experimental results prove that the method in this paper can ultimately obtain eight target-relevant color temperatures from 2 700 to 6 500 K by ANSI/NEMA C78.377-2017 standards, which are 2 700, 3 000, 3 500, 4 000, 4 500, 5 000, 5 700 and 6 500 K, and the spectra of the light sources under each CCT with their Color rendering index, color fidelity, color gamut index are above 90, color deviation value within 0.0054, blue light hazard radiation efficiency are within 0.1 of the hybrid LED white light spectrum.

In the practical high-temperature gas detection case, the measured results of the gas spectral line parameters are often affected by temperature changes. Sometimes, it is even difficult to achieve real-time online measurement.Therefore, this paper aims to design and process a new high-temperature sample cell to simulate a high-temperature environment and build a tunable diode laser absorption spectrum measurement system to detect the spectrums of the target gas in a high-temperature environment. In this way, the accurate detection of spectral line parameters can be achieved.In the design of a high-temperature sample cell, Comsol was used to simulate and analyze the solid thermal conductivity of various materials to determine the optimal processing materials and size. The results show that good properties of the high-temperature sample cell are obtained. It can work in the temperature range of 300~1 000 K and the pressure range of -0.1～10 atm. The maximum temperature deviation of the sample cell at 1 000 K is 20 K. The measured leakage rates at 300 and 1 000 K are 5 and 60 Pa·min-1, respectively.This paper uses a distributed feedback (DFB) semiconductor laser with a center wavelength of 1 573 nm as the light source to measure the partial high-temperature spectrums of CO molecules with relatively accurate parameters in the HITRAN2016 database. The comparison between the spectral line parameters obtained from the inversion and those in the HITRAN database indicates that the error is within 5%. The good performance of the designed high-temperature sample cell was proved, which can help in the measurement of gas spectral line parameters in high-temperature environments.

Impurity behavior study is very important in magnetic confined fusion research as impurity may cause the dilution of fuel ions, affect the power balance, and degrade plasma performance. Spectroscopic diagnostic is important for impurity measurement and transport study in fusion devices. Spectroscopy in the vacuum ultraviolet (VUV) range offers a useful tool for investigating impurity radiation from low-temperatureareasof edge tokamak plasma. To meet the impurity measurement requirement in fusion research, a Seya-Namioka spectrometer was designed, and the main parts of the spectrometer are adjustable width incident slits, concave gratings, and a detector). The luminous flux of the spectrometer is adjusted through an adjustable width of the incident slit, which is of the linear guide type. The position of the incident slit baffle is adjusted through the linear guide to achieve slit width adjustment. The grating surface is coated with aluminum (Al) and magnesium fluoride (MgF2) to enhance the refractive efficiency. The usable wavelength range of the grating is 50~460 nm, and the optical path optimization design is carried out for the 50~50 nm wavelength range. The spectrometers detector was chosen as a deeply cooled back-illuminated charge-coupled device (CCD). By turning the grating turntable to rotate the grating and change the diffraction angle, spectral observations in the 50~250 nm range can be achieved. Based on the parameters of the concave grating, the specific optical path was determined, and the relationship between wavelength and grating rotation angle, as well as the line dispersion rate at different wavelengths, was analyzed. According to the theory of concave grating imaging, the spectral resolution of the system was calculated and analyzed. The optimal exit arm was determined to be 205 mm by analyzing the spectral resolution under different exit arms. The effects of incident slit width, pixel size, aberration, and diffraction limit on spectral resolution were analyzed with an exit arm of 205 mm and an incident slit of 20 μm. The main contributions to the spectral resolution are the entrance slit width, pixel size, aberration, and diffraction limit. The diffraction limit has the smallest effect on spectral resolution, which can be ignored. Due to the size of the detector pixel, the width of the exit slit has a great impact on the spectral resolution, which remains at about 0.09~0.10 nm and increases slowly in the 50~250 nm wavelength range. The total spectral resolution is between 0.121 nm and 0.122 nm. The width of the incident slit will be adjustable from 10 to 1 000 μm, depending on the intensity of the incident light. The wider the width of the incident slit, the worse the overall spectral resolution will be. In actual measurement, the luminous flux and spectral resolution should be considered. Wavelength calibration and performance tests were performed by using a low-pressure mercury lamp and microwave plasma light source. The wavelength calibration of the spectrometer can be completed based on the zero-order spectrum and the characteristic wavelength of the mercury lamp (Hg Ⅰ 185 nm). The zero-order spectrum position is defined as the diffraction angle zero position. The angular position of Hg Ⅰ 185 nm in the spectrometer is determined according to the formula. The spectral resolution was 0.124 3 at 185 nm by Gauss fitting of the Hg Ⅰ (185 nm) spectral line, which is close to the calculated value. By comparing the spectral resolution of Hg Ⅰ 185 nm under different exit arms through experiments and theoretical calculations, it was further verified that the instrument achieved the best spectral resolution at the exit-arm length of 205 mm. Further performance tests of the spectrometer were carried out using a microwave plasma discharge light source device. It has been verified that the spectrometer has good detection ability in the 50~250 nm wavelength range, based on observing nitrogen, oxygen, and helium lines emitted by microwave plasma light sources.

In this study, Songyuan Grain Group Storages corn grain reserves from various years were used as samples, and a three-stage micro Raman spectrometer was used to gather the spectra of individual grains of corn from multiple conditions. The regularity of its internal components changes over time was investigated using the molecular spectra produced by Density Functional Theory (DFT). The DFT theory optimised the chemical structure, and the vibration frequency associated with each molecules bond was analyzed. The partial peak location was then determined by comparing the vibration frequency to the Raman spectra of corn grains as measured in practice. Strong Raman peaks were discovered in the original spectra after pretreatment at 476, 1 006, 1 156, 1 459, 1 519, 1 597, and 1 633 cm-1. It indicated the presence of lignin, lutein, and corn starch. Measurements were made of the Raman spectra of longitudinal corn sections from various years. After spectral normalisation, the relative peak intensities and areas of all peak sites were determined. In addition to the surface-based Raman peaks, fresh Raman peaks in the 850～1 450 cm-1 range had also been discovered. With more storage years, distinct changes occur in the starch contents Raman intensity, peak area, and FWHM at 475, 863, 944, 1 260, 1 338, and 1 378 cm-1. It demonstrates how its molecular structure, substance, and makeup have distinctly altered. This modification essentially fits the corn starch aging process. The Raman peaks at 1 080 and 1 130 cm-1 revealed a significant variation in the vertical section Raman spectra of maize across three years. While the Raman peak at 1 192 cm-1 only occurs in the spectrum of maize stored in 2020, the peak strength and FWHM value varies to variable degrees, the Raman peak at 1 080 cm-1 is only present in the spectrum of corn stored in 2018 and 2019. It suggests that the reciprocal conversion of specific carbohydrate compounds in starch can be utilized as a distinctive peak to distinguish between corn samples from various years. The peak strength and peak area values were assessed by linear fitting with the year after the physical properties of the peak sites associated withlutein content were compared. The coefficients were above 0.9, and the fitting degree was good, showing that the lutein concentration declined linearly as storage duration increased. The loss of lutein has varied impacts on the vibration of chemical bonds, with the stretching vibration of the CC double bond being the most affected. It is discovered by comparing and evaluating the spectrum of lutein acquired from theoretical calculation. This study investigates the changes in the Raman spectra of corn with time using density functional theory calculations and experimental comparisons. It gives a definite basis for identifying corn years using Raman spectroscopy. It increases the application of spectroscopy in agricultural product quality detection and analysis by analysing the internal composition variations of maize in different years.

The chlorophyll and carotenoid content is an important indicator for evaluating the health status of plants. The PROSPECT model, coupled with machine learning, has been widely used to retrieve the biochemical properties of vegetation. However, the application of the coupled model is limited due to the differences between the leaf-directional hemispherical reflectance factor (DHRF) spectra and the bidirectional reflectance factor (BRF) spectra. This paper utilizes the leaf spectral database of North American plant (EcoSIS) as the experimental dataset and introduces the PROSPECT model as an additional constraint for machine learning. This approach creates a hybrid dataset by employing wavelet continuous wavelet transform (CWT) to generate the wavelet coefficient spectrum and the derivative spectrum generated by the first-order derivative (FD). Three kinds of feature extraction algorithms, namely competitive adaptive reweighting algorithm (CARS), successive projection algorithm (SPA), and principal component analysis (PCA) were applied to extract spectral features for chlorophylls and carotenoids in the full-spectral domains and the subdomain of VNIR spectroscopy. Based on the above 12 combinations of different methods, artificial neural network (ANN) prediction models for chlorophyll and carotenoids were separately established. The results show that the simulated data under the constraint of the PROSPECT model enhanced the quality of the training set for machine learning to a certain extent. Additionally, the spectra processed by the first-order derivatives and wavelet transforms were able to reduce better the bias between the simulated spectra of the DHRF and the measured spectra of the BRF. The best inversion of leaf chlorophyll is achieved with the FD+CARS combination in the whole spectral domain, yielding a test set R2 of 0.806 4 and RMSE of 2.911 4. Meanwhile, the CWT+CARS combination in the VNIR spectral sub-domain offers the best results for leaf carotenoids, with a test set R2 of 0.797 2 and RMSE of 0.414 1. The proposed method can provide researchers with a reference to extract biochemical characteristics of plant leaves more accurately and efficiently from BRF spectra and other near-end reflectance images.

The complexity of the meat processing process presents significant challenges in detecting adulteration in meat products. This study uses hyperspectral technology to identify and analyze adulteration in beef meatballs. To establish the models, different proportions (20%, 40%, and 80%) of pork/chicken were added to mince beef to obtain single adulterated samples, respectively. Subsequently, pork and chicken were mixed in 2∶8, 5∶5, and 8∶2 ratios to prepare samples for composite adulteration under three different gradients (20%, 40%, and 80%). In addition, fried adulterated beef balls were also prepared to test the applicability of classification models. Hyperspectral data of the adulterated samples were collected and preprocessed using five different methods. Adulteration identification models were developed using the Extreme Learning Machine Classification (ELMC) and Support Vector Classification (SVC) algorithms. Feature wavelengths were extracted using the Successive Projections Algorithm (SPA) and Competitive Adaptive Reweighted Sampling (CARS), developing corresponding simplified models. The results showed that the performance of the raw/cooked beef ball classification models SVC model based on full wavelength was better than that of ELMC. In contrast, the simplified model based on characteristic wavelength showed a contrary trend. For the classification of raw beef balls, the ELMC model (SPA-ELMC-Raw) established based on the 44 characteristic wavelengths selected by SPA yielded the best performance, with classification accuracies of 97.17% for both the calibration set and prediction set. For the classification of cooked beef balls, the ELMC model (CARS-ELMC-Cooked) established based on the 38 characteristic wavelengths selected by CARS showed the highest performance, with classification accuracies of 97.17% and 96.23% for the calibration set and prediction set, respectively. The results indicated that hyperspectral imaging technology proves to be an effective, rapid, and accurate method for discriminating between different types of adulteration in raw and cooked meat. This provides a strong theoretical basis for developing portable detection equipment.

The oil-in-water emulsion of sea oil spills has caused serious harm to the ocean. The identification and quantitative analysis of oil-in-water emulsion is of great significance for treating marine pollution and restoring the marine environment. In recent years, some studies have conducted quantitative analysis on thinner emulsified oil spills, but there are rare in-depth studies on thicker emulsified oil spills. The fluorescence produced by the oil-in-water emulsion irradiated by the laser on the sea surface will begin to appear saturated under certain conditions and will not change anymore. This paper studies the spectral characteristics of the oil-in-water emulsion at this critical point of saturation and related ranges. According to the relationship between the optical power received by the LIF detection system and the BRRDF value, the fluorescence spectrum is simulated using the BRRDF model. The analysis uses Romashkino oil and Petrobaltic oil, representing dark opaque and bright transparent oil.A method for calculating the fluorescence intensity of thicker oil-in-water emulsions was proposed, and the thickness at which fluorescence saturation appeared in emulsions of two oils at different concentrations and emulsification times was calculated. Comparing the calculated results, it can be seen that the saturated thickness decreases with the increase of concentration and decreases with the increase of emulsification time. The analysis results show that for the oil-in-water emulsion of dark opaque oil in seawater, the fluorescence intensity received by LIF is usually saturated, so the thickness of the emulsion layer exceeding the saturation thickness cannot be judged by using the LIF system alone. A neural network with 4 layers of neurons was designed to verify the ability of the fluorescence spectrum near the saturation critical point to identify the emulsion concentration. The verification results show that as long as the oil-in-water emulsions of different concentrations reach or exceed the saturation thickness, the fluorescence spectrum has good discrimination ability, which can be used to distinguish the emulsions of different concentrations. For those samples far from the saturation thickness, fluorescence spectroscopy can distinguish whether the emulsion has reached saturation thickness. These experiments and conclusions will provide a reference for identifying and quantitatively analyzing thicker oil-in-water emulsions.

When X-ray fluorescence (XRF) is used in analyzing the element composition of the sample in air atmosphere, the strong absorption from the air to the low-energy characteristic X-rays from such low atomic number (Z) elements as Cl, K, Ca, and so on, deeply affect the analysis for them.To avoid the strong absorption from the air to the characteristic X-ray of low Z elements, the XRF technology in a vacuum atmosphere is often used, but it requires a complicated and expensive vacuum system. Besides, the power of the laboratory source is generally low. This results in a low intensity of incident X-rays, which also affects the analysis of the element composition of the sample with XRF. To solve the above problems, a simple closed XRF analysis system in a helium atmosphere based on a monolithic polycapillary X-ray lens (PCXRL) and rotating target X-ray source was designed. Strong XRF signals were obtained by using X-rays with a high gain of power in density focused by the PCXRL to irradiate the sample, and the excitation channel and the detection channel were both in a stable helium atmosphere to reduce the absorption from the air to the characteristic X-rays of low Z elements. The designed XRF system was characterized to show that with a rotating molybdenum target working at a voltage of 29 kV and a current of 20 mA, the detected XRF intensity of Cl, K, Ca, and Fe in a helium atmosphere is higher than that in an air atmosphere, respectively. For elements with an energy of the characteristic XRF below 8 keV in plants, the characteristic XRF intensity detected in the helium atmosphere is 1.1 to 5.5 times that in the air. This is helpful for an efficient and non-destructive XRF analysis of the elements with low Z of samples.

Near-infrared (NIR) spectroscopy detection technology can reflect the measured samples organic chemical composition and structural information by detecting the spectral features in the NIR region. During the material composition analysis, NIR spectroscopy often involves a significant amount of wavelength data, resulting in relatively high data dimensions. Furthermore, spectra are susceptible to phenomena such as overlap and redundancy, which impact the models performance. Therefore, we proposed a noise discriminant C-means clustering (NDCM) algorithm that combined fast generalized noise clustering (FGNC) and fuzzy linear discrimination analysis (FLDA). NDCM can realize the extraction of data identification information and data space compression in the fuzzy clustering process, which can achieve higher clustering accuracy. The fuzzy membership degree and the cluster centers obtained by fuzzy C-means clustering (FCM) on the near-infrared spectral data of Chuzhou chrysanthemum tea are used as the initial fuzzy membership degree and initial clustering centers of NDCM, respectively, so that NDCM has the advantages of fast clustering speed and high accuracy. The FCM algorithm is sensitive to noisy data, while the NDCM algorithm can perform better when dealing with noisy data in spectra. In this study, 240 samples of Chuzhou chrysanthemum tea with three quality grades, namely special grade, first grade and second grade, were selected as experimental samples. A portable NIR spectrometer (NIR-M-F1-C) was used to collect the NIR spectra of Chuzhou chrysanthemum tea, and they are the 400-dimensional data. At first, the NIR spectra were pretreated with Savitzky-Golay filtering and multivariate scattering correction (MSC) to reduce spectral scattering and noise. Secondly, the dimensionality of the spectral data was reduced by principal component analysis (PCA), and the dimensionality of the data after PCA reduction was 6. Next, linear discriminant analysis (LDA) was applied to extract the discriminant information in the spectral data of Chuzhou Chrysanthemum tea and further transform the data space dimension into 2 dimensions. Finally, three algorithms, i.e. FCM, FGNC and NDCM, were utilized to perform cluster analysis on the processed data to accurately classify chrysanthemum tea. The experimental results exhibited that when the weight index m=2.5, the clustering accuracy rates of FCM, FGNC and NDCM were 92.42%, 98.48%, and 100%, respectively. The clustering time of NDCM was slightly longer compared to FGNC. FCM had 27 iterations to reach convergence, while FGNC and NDCM took 13 and 10 times, respectively. NIR spectroscopy combined with MSC, Savitzky-Golay filtering, PCA, LDA and NDCM can provide a clustering model to accurately identify Chuzhou chrysanthemum tea quality.

Leaf nitrogen content (LNC) is crucial for assessing plant growth status and photosynthetic capacity. Accurate LNC can aid in the rational control of nitrogen fertilizer application, which is critical for achieving efficient agricultural production. Chemical analysis methods can accurately detect nitrogen content. However, it often requires destructive sampling and cumbersome steps, which are difficult to measure in real-time. Spectral technology can enable nondestructive detection of LNC, but the high dimensionality and noise inherent in spectral data make accurate estimation challenging for precision agriculture. To achieve accurate prediction of nitrogen content in eggplant leaves, this paper proposed a feature extraction method of spectral data based on hyperspectral imaging (HSI) technology and a one-dimensional convolutional autoencoder network (CAE). The proposed method utilized pixel-level spectral data to train the CAE, fully utilizing the HSI data of leaves. This can extract deep features that retain local spectral features related to the nitrogen content distribution on the leaf surface, reducing data dimension, filtering out noise, and enhancing the accuracy and stability of the nitrogen content prediction model. In this paper, we set up four nitrogen application gradients for eggplant, obtained leaf samples with varying nitrogen content using culture, and measured their HIS data. Multiple scattering correction algorithm was used for data preprocessing. The HSI-CAE method, competitive adaptive reweighting (CARS) algorithm, and random frog (RF) algorithm were used to extract spectral datas deep features and characteristic wavelengths, respectively. The partial least squares regression(PLSR)models were built based on these features. The influence of deep features and different feature wavelength combinations on the accuracy of the prediction model was compared to determine the optimal feature extraction method. The results were as follows: the test set determination coefficient of the prediction model, established by using deep features from the CAE encoder of different depths, was greater than 0.85. When 28-dimensional features were output, the test set determination coefficient was 0.910 2, and the root mean square error was 3.118 9 mg·g-1. It was found that the CAE-PLSR model has the best prediction performance, which verified the feasibility and superiority of the HSI-CAE feature extraction method for estimating nitrogen content in eggplant leaves. In conclusion, the HSI-CAE feature extraction method can efficiently analyze HSI data and extract its deep features. These features contain information highly related to nitrogen content. The deep feature modeling used in this research greatly reduced the complexity of the model. It effectively improved the accuracy of the nitrogen content prediction model, providing a new way of implementing accurate prediction of nitrogen content based on the HSI technology.

Lutein is a natural antioxidant that has numerous benefits for human health. Heterotrophic Chlorella sorokiniana has the advantage of high purity and production of lutein. In contrast, the production of lutein in Chlorella sorokiniana mainly depends on two factors: biomass productivity and lutein content. However, conventional approaches such as the optical density method for measuring biomass productivity and high-performance liquid chromatography for measuring lutein content suffer from drawbacks, including complex procedures and limited timeliness. A visible near-infrared dual-mode snapshot multispectral imaging detection system was constructed to rapidly and non-destructively determine the variations in lutein production during the growth process of Chlorella sorokiniana. Based on the spectral response range, the visible camera was used to obtain the spectral information image of lutein content, and the near-infrared camera was used to obtain the spectral information image of biomass productivity to build a visible near-infrared dual mode multispectral dataset containing biomass productivity and lutein content information. To address the issue of wide spectral range and limited wavelengths in the snapshot multispectral camera used in the system, a novel approach combining sequential floating forward selection with a modified successive projections algorithm (mSPA) was proposed. A comparative study was conducted, evaluating mSPA against successive projections algorithm, genetic algorithm, and random frog algorithm for wavelength selection. Multiple linear regression and extreme learning machine models were constructed based on the selected feature wavelengths. Finally, the optimal predictive models for biomass productivity and lutein content were used to generate a visualization distribution map of lutein production in Chlorella sorokiniana. The results indicated that when using near-infrared and visible cameras for biomass productivity and lutein detection in Chlorella sorokiniana, the mSPA algorithm consistently yielded fewer feature wavelengths for both biomass productivity and lutein and achieved the highest prediction accuracy. The optimal models of biomass productivity and lutein content were established using the mSPA-selected feature wavelengths in combination with an extreme learning machine. The corresponding coefficients of determination for the prediction sets were 0.947 for biomass productivity and 0.907 for lutein, with root mean square errors of 0.698 g·L-1 and 0.077 mg·g-1 and residual prediction deviations of 3.535 and 3.338, respectively. The models demonstrated good predictive capabilities. The visualization distribution successfully achieved intuitive monitoring of lutein production variations in Chlorella sorokiniana, which is beneficial for online detection of lutein content in practical production scenarios. The mSPA algorithm, employed in the snapshot multispectral detection of biomass productivity and lutein content in Chlorella sorokiniana, effectively avoided the incorrect selection and omission of feature wavelengths by evaluating each sorted wavelength individually, thereby improving the prediction accuracy of the models. This approach provides a new wavelength selection strategy for applying snapshot multispectral imaging technology.

Sequential Batch Reactor (SBR) is one of the most widely used active sludge treatment devices.In this experiment, the nitrate and nitrite nitrogen content changes during the startup process of the short-term nitrification reaction system were studied using an SBR reactor to treat artificially simulated high ammonia wastewater.A model was established based on the data collected usingultraviolet spectroscopy,aiming to rapidly predict the nitrate nitrogen and nitrite nitrogen content in the effluent of the SBR reactor. Using laboratory-prepared solutions with different concentrations of nitrate and nitrite nitrogen, a calibration model for standard mixtures was constructedusing the interval partial least squares(iPLS) for three different band selection methods. The research results show that the models built exhibit good correlations between the measured and predicted values for nitrate nitrogen and nitrite nitrogen in the mixed solution.To determine the reactor effluent parameters, models for ultraviolet spectroscopy and the nitrate and nitrite nitrogen content were constructed using partial least squares algorithms for three different band selections.The model results were evaluated using the calibration set correlation coefficient, the root mean square error of cross-validation (RMSECV), the correlation coefficient of the prediction set, and the root mean square error of prediction (RMSEP) evaluation metrics.Among the three models, the model built using the synergy interval partial least squares (siPLS) method, which divided the full spectrum into 24 and 19 intervals and established models for the combined sub-intervals［2 4］ and ［3 8］, exhibited the best prediction and fitting results. Its calibration model hadr=0.939 3 and RMSECV=1.650 4, and the prediction model hadr=0.950 7 and RMSEP=0.442 1. This model showed an overall good prediction performance for nitrate nitrogen and nitrite nitrogen, indicating that establishing interval partial least squares models using ultraviolet spectroscopy can rapidly predict nitrate nitrogen and nitrite nitrogen content in the effluent of the short-term nitrification reactor.

The oil leakage fault detection of traction transformers isa crucial component of Electrical Multiple Unit (EMU) operation fault detection. In the near-infrared image of the bottom plate collected by the trouble of moving the EMU detection system, the characteristics of leakage transformer oil traces are similar to those of water traces, which cannot be distinguished by the trouble inspectors and image recognition algorithms, resulting in a high false alarm rate of the oil leakage fault and affecting the operation and maintenance efficiency of the train.Water lacks fluorescence properties, whereas mineral transformer oil exhibits ultraviolet fluorescence, so fluorescence imaging can differentiate the two traces. Moreover, varying operating seasons, regions, and periodsof EMU expose the leaked transformer oil to diverse environmental conditions, affecting its fluorescence properties. The fluorescence properties of the leaked oil decide the choice of excitation light source wavelength and imaging band for the fluorescence imaging system.Hence, samples under different temperatures and light conditions are first prepared to explore the fluorescence properties of mineral transformer oil used by EMU and the impact of two major environmental factors, temperature, and light, on its fluorescence properties. Then, the three-dimensional fluorescence spectra of the samples were collected by the fluorescence spectrometer and subsequently analyzed, the fluorescence characteristics of the samples and the effect of temperature and illumination on the location and intensity of fluorescence characteristic peaks. The results show that the mineral transformer oil for EMU has two strong fluorescence excitation/emission peaks located at 350/382 and 350/402 nm, respectively. The increase in temperature causes a decrease in fluorescence intensity but does not change the position of the fluorescence peaks. The increase of light intensity and the extension of light duration cause a decrease in fluorescence intensity but do not significantly change the position of fluorescence peaks. The experimental results provide experimental evidence and data support for fluorescence imaging to detect the leakage fault of mineral transformer oil for EMU.

Octadecylamine is a flotation collector commonly used to produce potassium chloride by cold crystallization-positive flotation, and its dosage will significantly affect the flotation separation efficiency. In addition, octadecylamine will adsorb on the surface of potassium chloride and inevitably remain in potassium chloride products, which is inconducive to developing high-purity potassium salt products. To meet the requirement of determination of octadecamide potency in potassium chloride, an extraction spectrophotometry was developed, based on the principle of van der Waals force and hydrogen bonding between octadecylamine and bromophenol blue with butyl acetate as extractant and bromophenol blue sodium salt as chromogenic agent. The effects of solution pH, dodecyl morpholine, co-existing salt, equilibrium time, and the amount of chromogenic agent on the concentration measurement of octachylamine were investigated. The results show that when the solution pH increases from 3 to 9, the absorbance of the extract decreases because the complexation between octachylamine and bromophenol blue weakens. The polarity of the complex molecules increases. When the solution pH is less than 5, dodecyl morpholine and bromophenol blue can also form colored complexes. While pH is between 6 and 9, the absorbance tends to zero. The ionic strength of the coexisting solutions of potassium chloride, sodium chloride, potassium sulfate and magnesium chloride increases, weakening the hydrogen bonding. Still, the salting-out effect of the complex molecules is enhanced, and the trace Li+, NH+4 and B in the solution have little effect on the absorbance. The absorbance does not conform to Lambert-Beer law when the amount of chromogenic agent is too large. Besides, equilibrium time has little effect on the absorbance. The determination conditions in potassium chloride solution are as follows: The solution pH is 6, the ionic strength is 1 mol·L-1, and the amount of 2 mmol·L-1 chromogenic agent is 0.5 mL. 5 mL butyl acetate is added to the 25 mL aqueous solution adjusted by buffer after 5 min of reaction for extraction. After 2 min of delamination equilibrium, the absorbance of the extraction solution is tested at 458 nm. The working curve is A=0.049 49c+0.066 24 (R2=0.992 3, ε=1.33×104 L·mol-1·cm-1, 0~10 mg·L-1). The relative standard deviation of this method was 0.33%~2.63%, the mean relative error was -0.90%, and the mean relative error of the system of octadecylamine and dodecylmorpholine was -0.25%. The content of octadecamide in the filter liquor derived from washing potassium chloride after positive flotation was 8.66 mg·L-1 and recoveres were 95.5%~106% by this method. The extraction spectrophotometry has been proven to be appropriate for detecting the concentration of octadecamide in the production of potassic fertilizer in salt lakes.

Chromium has multiple roles in nickel-based alloy and has outstanding contributions to increasing the wear resistance of alloy coating. At the same time, chromium can form a dense oxidation protective film under the action of high-temperature gas, which significantly improves the alloys resistance to high-temperature oxidation and thermal fatigue performance. However, in the case of high chromium content, chromium easily forms harmful phases with titanium, aluminum, molybdenum, and other elements in the alloy and reduces its strength. So, the quasi-determined value of chromium in nickel-base alloy is very important. Nickel base alloy samples are digested by microwave digestion using hydrofluoric acid, hydrochloric acid, and nitric acid. The matrix is matched by inductively coupled plasma atomic emission spectrometry, and the calibration solution is prepared to determine-chromiums net strength in the nickel-base alloys analysis spectrum line to calculate the content. Given the interference of high-content elements such as Ni, Mo, and Fe in nickel-base alloys on the determination of Cr content, 266.602 nm with fewer interference elements and moderate sensitivity was selected as the analytical spectral line; the matrix effect and the interference effect of coexisting elements were systematically investigated. When preparing the calibration solution, the elements with content greater than 5% were matched with the same content. Prepare the low standard calibration solution Kl Crfor the chromium element, prepare the high standard calibration solution Kh Crfor the chromium element, prepare the sample solution KS Cr, and use the ICP-OES spectrometer to determine the net strength of the low standard calibration solution Kl Cr, the sample solution KS Cr, and the high standard calibration solution Kh Cron the analytical spectrum line in turn, and calculate the chromium content according to the formula. The methods accuracy was evaluated using standard reference materials, and the reliability of the method was verified by comparative analysis using the YS/T 539.4-2009 standard. The results showed that the limit of detection (LOD) and limit of quantification (LOQ) of the method were 0.05 and 0.10 μg·mL-1, respectively, and the relative standard deviation (RSD) of the determination result is less than 2.5%. The relative error (RE) of the comparison analysis result between the measured value and the certified value of the standard reference material is less than 2.5%. This method is easy to operate and accurate to detect. It broadens the detection range for the determination of chromium content in nickel-based alloys by ICP-OES. The lower detection limit is reduced from 2.0% to 0.1%, and the upper limit of detection range is increased from 30% to 33.5%. It greatly improves the efficiency and accuracy of the detection of chromium content in nickel-base alloys. It is suitable for determining chromium content in various brands of nickel-base alloys.

Fourier transform infrared spectroscopy is a commonly used in-situ non-destructive analysis technique for qualitatively identifying natural fluid inclusion components. The quantitative analysis of temperature-pressure components of natural inclusions is the focus, as well as the difficulty and development direction of laser spectroscopy geological applications. The quantitative method and Fourier transform infrared spectroscopy model for natural fluid inclusions have not been established. The infrared spectra of a single CO2 system with two stretching vibration peaks of 2ν2+ν3 and ν1+ν3 were collected and observed at 40～80 ℃ and 60～500 bar for the first time by high-pressure capillary silicon tube packaging technology combined with Fourier infrared spectrometer. The infrared spectral parameters at different temperature and pressure ranges were obtained. The peak area and peak displacement of the Fourier infrared spectrum under this system were obtained by fitting and calculation. The calibration model of the density and peak area, pressure, and peak area of the Fourier infrared spectrum fluid under the 2ν2+ν3 and ν1+ν3 stretching vibration bimodal system of CO2 was established. The variation characteristics of bimodal stretching vibration of 2ν2+ν3 and ν1+ν3 of CO2 with temperature and pressure are clarified, and the variation of peak displacement with temperature and pressure is discussed. Using Fourier transform infrared spectroscopy (FT-IR) of a single CO2 component, the densities of pure CO2 inclusions in quartz fractures of the Huangliu Formation reservoir in the Yinggehai Basin were calculated at 40 ℃ by collecting the double peak area of stretching vibration of the inclusions. At the same time, the Raman spectra of the inclusions were collected at room temperature, and the densities calculated by two different in-situ analysis methods were compared to verify the feasibility of FT-IR quantitative analysis. Compared with Raman quantitative analysis, Fourier transform infrared spectroscopy quantitative analysis is not limited by experimental instruments and environment. Its quantitative method can avoid errors caused by different laboratories where the instrument is located.

In recent years, the rapid development of modern spectral detection technology is closely related to deep learning. As an end-to-end model, the deep neural network can get more information from the spectra, thus improving the robustness of the model. A one-dimensional residual neural network (1D-AD-ResNet-18) model based on a convolutional block attention module was proposed to explore the feasibility of qualitative prediction of mango species by near-infrared spectroscopy combined with deep learning. Firstly, to reduce the interference of redundant information in the spectra, the CBAM convolution attention module is added to the traditional one-dimensional residual neural network, which can focus on the local useful information of the spectra. Secondly, to avoid the disappearance of gradient and the occurrence of overfitting, ResNet-18 is used to solve the problem of network “degradation”. For 186 mango samples, 70% of the samples were trained, and 30% were tested. Accuracy, Precision, Recall, F1-score, Macro-average, and weighted average were used as evaluation indexes of the model. Three comparison models were established, including traditional one-dimensional ResNet-18, SNV-SVM, and PCA-KNN. Compared with the above three methods, the established 1D-AD-ResNet-18 model obtained the optimal prediction results, and the accuracy of the four qualitative analysis models was 96.42%，80.35%，76.78% and 67.85%. The experimental results show that the 1D-AD-ResNet-18 model can accurately identify and classify mango species, which provides a new idea for the qualitative analysis of mango species by NIR spectroscopy.

Lettuce is one of the vegetables that people often eat, and the storage time of lettuce is an important factor affecting the freshness of lettuce. Therefore, it is necessary to develop a simple, fast, and non-destructive method to identify the storage time of lettuce. Near-infrared spectroscopy (NIR) can quickly and accurately detect the near-infrared spectrum of lettuce to realize the non-destructive identification of lettuce storage time. However, noise and redundant signals are in the NIR spectral data collected by the near-infrared spectrometer. To eliminate the noise information of the spectrum and extract the feature information, a novel method was proposed to identify the storage time of lettuce based on NIR spectroscopy and fuzzy uncorrelated QR analysis (FUQRA). Firstly, principal component analysis (PCA) was used to reduce the dimension of the original spectral data from 1557 dimensions to 22 dimensions. Secondly, after the feature vectors are obtained by fuzzy uncorrelated discriminant transformation (FUDT), the discriminant vector matrix is established by using the feature vectors, and the final discriminant vector matrix is obtained by QR decomposition. Finally, the k-nearest neighbor algorithm was utilized for classification. 60 fresh lettuce samples were selected as the research object. Firstly, the NIR spectral data of lettuce samples were collected by Antaris Ⅱ near-infrared spectrometer and detected once every 12 hours. Secondly, multivariate scatter correction (MSC) was used to reduce the noise signal in the NIR spectra. To verify the effectiveness of the proposed method, the experimental results were compared by four classification models: principal component analysis (PCA) combined with a K-nearest neighbor (KNN) algorithm, PCA and fuzzy linear discriminant analysis (FLDA) combined with KNN algorithm, PCA and fuzzy uncorrelated discriminant transformation (FUDT) combined with KNN algorithm and PCA and FUQRA combined with KNN algorithm. The classification accuracies produced by different values of the weight indexm were studied, and the most appropriate parameters were selected: m=2, K=3. Under this condition, the classification accuracies of the four algorithms were 43.33%, 96.67%, 96.67%, and 98.33%, respectively. It can be seen that compared with the other three algorithms, FUQRA can better realize the identification of lettuce storage time.

The emergence of the novel coronavirus has caused significant losses in the global economy and public safety. The need for efficient detection and diagnosis has become urgent as the virus continues to inflict damage worldwide. Accurate and fast diagnosis of the novel coronavirus is critical for epidemic prevention and control. With the development of technology, deep learning has made many breakthroughs in detection and recognition, attracting widespread attention from researchers. However, deep learning requires a large amount of data for model training, and the collection of Raman spectra data for the novel coronavirus is limited by devices and environmental factors, making it difficult to obtain large amounts of data. The limited training data can hinder the training of deep learning models and limit their accuracy, resulting in poor performance in actual detection. To address this problem, this study introduces the CGAN adversarial network to automatically extract features from Raman spectra data and generate new spectra to expand the dataset for the novel coronavirus. These methods can effectively increase the training sets size and improve the models accuracy. Using the enhanced Raman spectra dataset and deep neural networks, as well as traditional machine learning methods, including logistic regression, decision trees, random forests, support vector machines, and k-nearest neighbors algorithm, the diagnosis of COVID-19 is performed. The experimental results show that the deep neural network has a prediction accuracy of 98% for whether a patient is infected with the novel coronavirus, which is higher than traditional machine learning algorithms, demonstrating the superiority of deep learning models in Raman spectra detection of the novel coronavirus. We also compared the performance of the datasets before and after augmentation in different models, proving the effectiveness of data augmentation. Compared with traditional detection methods, this method is non-invasive, fast, and accurate, providing a new approach for biomedical detection of the novel coronavirus. The method proposed in this study can assist in rapidly and accurately detecting the novel coronavirus. It can be applied not only to the diagnosis of the novel coronavirus but also to the diagnosis of other diseases, having practical application value.

This study focuses on identifying and detecting spectral molecular bands caused by changes in internal energy levels of molecules, which contributes to the research of stellar spectral types and parameter estimation. First, considering the curve trend of molecular bands, pseudo molecular bands that have a W shape but an obvious downward trend should be eliminated by using curve analysis to identify molecular bands. Bringing the identification idea of multi-type and multi-classification criteriainto the model, the four parameters of the detected molecular band peak depth, W-shaped width, curve trend, and rebound trend are adoptedas training features, which consider comprehensivelythe change rate of starting point, change trend of curve, extreme point distribution and the factors of curve shape. Secondly, the LightGBM (Light Gradient Boosting Machine) model is used to identify the spectral and molecular band characteristic parameters of F-type stars with an accuracy of 97.62% and 99.16%, respectively, to verify the feasibility and reliability of this method. This work can not only excavate the late stars and improve the accuracy of data labels but also automatically identify the late stars by using the LightGBM machine learning model to detect the unknown spectrum based on accurate recognition, which improves recognition efficiency and reduces memory occupation.

In recent years, there has been a significant increase in the number of observed astronomical spectra. This has led to greater demands for automatic classification and analysis of spectra in large-scale spectroscopic surveys. In this study, we introduce a multi-scale feature fusion-based stellar spectral classification model (MSFnet) that leverages the strengths of Convolutional Neural Networks (CNNs) in classification tasks and employs a multi-scale feature fusion module to extract spectral features at various scales for the prediction of stellar spectral types. The proposed MSFnet architecture consists primarily of a multi-scale feature fusion module and a CNN with four convolutional layers, two max-pooling layers, and one fully connected layer. To mitigate overfitting, dropout is incorporated into the model, enhancing its robustness by reducing dependence on specific local features. The dataset employed in this study is derived from the LAMOST DR9 database. Before training, data preprocessing is performed, which includes uniform resampling of spectra and min-max normalization. The experiment uses Python 3.9 and the PyTorch deep learning framework to build the network. The experimental section is divided into two parts: the first part investigates the relationship between the number of layers in the CNN, the number of feature maps, and the classification accuracy; the second part compares the performance of the proposed MSFnet model and the Resnet18 model using evaluation metrics such as precision (P), recall (R), and F1 score. Both models training, validation, and test sets are split according to a 6∶2∶2 ratio to maintain consistency in training samples. Results demonstrate that the highest accuracy is achieved using a CNN with four convolutional layers and 16 feature maps. Based on this finding, we propose the MSFnet model, which combines the feature fusion module with the CNN. Compared to the 18-layer residual neural network model, the MSFnet model has a more straight forward structure and performs similarly on the evaluation metrics. The performance in the metrics above is on par with that of the Resnet18 model. Furthermore, it demonstrates superior classification efficacy for spectra types A, F, and K, accompanied by enhanced speed. With an accuracy of nearly 97% on the test set, the MSFnet model outperforms traditional CNN and Resnet18 models, indicating its potential to improve the accuracy of automatic spectral classification significantly.

Since their excavation nearly 70 years ago, the jade artifacts from the Beiyinyangying site have been lacking in scientific testing and analysis. To address this issue, 13 jade artifacts from the Beiyinyangying site in the Shanghai Museum collection underwent non-destructive testing and analysis using a super-depth-of-field optical microscope, Fourier transform infrared spectroscopy (FTIR), portable X-ray fluorescence spectrometry (pXRF), and Raman spectroscopy. The spectroscopic results confirmed that the jade artifacts were made of micaceous jade, quartzose jade, and idocrase jade. This finding, combined with the previous visual identification results by scholars, reinforces the understanding that nephrite was not widely used during the Beiyinyangying cultural period. Additionally, Raman spectroscopic analysis revealed the presence of special mineral inclusions, such as rutile, monazite, and strengite, in the micaceous jade from the Beiyinyangying site. The jade exhibits high aluminum and low iron geochemical characteristics, indicating that it is a hydrothermal alteration product of intermediate-acidic volcanic rocks, which formed in a high-temperature, Ti-rich geochemical environment. Given the extensive distribution of Mesozoic volcanic and subvolcanic rocks in the lower reaches of the Yangtze River region, there is a possibility of the formation of medium-low temperature hydrothermally altered micaceous rocks. Moreover, considering the archaeological findings of micaceous jade artifacts at multiple prehistoric sites, it is assumed that the raw materials were likely sourced locally. However, further scientific research is necessary to draw a definitive conclusion.

In order to predict the water content and chlorophyll content of cantaloupe leaves quickly and accurately and improve the accurate management level of Cantaloupe crops, the leaves of cantaloupe in three different growth stages, namely the growing stage, the flowering stage, and the fruiting stage, were selected as experimental research objects by using the spectrophotometry technology. The correlation changes of leaf temperature, leaf water content, and chlorophyll content with LAB eigenvalues of color space were studied in three different collection periods: 9:00—10:00, 14:00—15:00, and 20:00—21:00, respectively. The least square method (LS) was used to preprocess the changes in temperature, water content, chlorophyll content, and color eigenvalues of different samples, and the eigenvalues with the best fit were selected for regression analysis and prediction model verification. The results showed that① Leaf temperature, leaf water content, and chlorophyll content had different color eigenvalues under different parameters. ② For leaves with 84%~93% moisture content, leaf temperature, and chlorophyll content were negatively correlated with leaf moisture content. ③ the chlorophyll content and leaf water content and color space LAB, there is a linear correlation. As the leaf water content -rises, L is on the rise, and the color becomes shallow gradually with light green leaves; with the increase of chlorophyll, L has a downward trend, showing the leaf color gradually deepens with black-green, exists in all types of sample data, L B positive. ④ Through model prediction and evaluation, random forest (RF), partial least squares (PLS), support vector machine (SVM), and LASSO can be used to predict chlorophyll content effectively. Among the chlorophyll prediction models, RF had the best prediction performance, R2c=0.939, RMSEC=0.868 and MAE=0.686, R2p=0.915, RMSEP=1.194 and MAE=0.942. ⑤ Through model prediction and evaluation, RF, PLS, AdaBoost, and polynomial regression (POLYNOMIAL) can effectivelypredict leaf water contents. In the prediction model of leaf moisture content, the POLYNOMIAL prediction performance is the best, R2c=0.884, RMSEC=0.005 9 and MAE=0.005 2, R2p=0.920 and RMSEP=0.006 2 and MAE=0.005 7. The spectrophotometry method can effectively and rapidly determine leaf water and chlorophyll content, which is expected to provide an optional feasible method for nondestructive, rapid, and accurate determination of leaf water and chlorophyll content.

Convolutional Neural Network (CNN) has a great advantage in data feature extraction, as it can fully acquire data features and has better generalization than traditional models. This study used a hyperspectral prediction method and modeling of Soil Organic Matter (SOM) content based on CNN. Using 320 soil samples from Shangzhuang Experimental Station, Changping District, Beijing, 807 spectral bands within 350~1 700 nm in the visible-near-infrared (VIS-NIR) were extracted, and the spectral data were denoised and transformed by the multivariate scattering correction (MSC) and the first-order differential transform. Successive projection algorithm (SPA) and competitive adaptive reweighted Sampling (CARS) were used to screen the sensitive wavelengths to realize the dimensionality reduction of the spectral data, respectively. To solve the problems of poor generalization of traditional means as well as the complexity and overload of deep CNN networks, based on the CARS and SPA algorithms, a shallow CNN model prediction based on 6 convolutional layers is proposed, and 1D-CNN1 and 1D-CNN2 with different convolutional sizes and number of convolutions are compared to find the optimal network parameters. By comparing the performance of VGG16, Support Vector Regression (SVR), Partial Least Squares Regression (PLSR), and Random Forests (RF) to build a prediction model in the feature wavelength and the full waveform. The optimal model was determined. The results show that compared with the full-spectrum band and SPA filtering algorithms, the model based on CARS filtering feature wavelength modeling performs better, and the number of bands is compressed to 8% of the full-wavelength band, which effectively realizes the dimensionality reduction of the spectral data. Comparing the full-band data, 1D-CNN1 and 1D-CNN2 based on CARS screening wavelengths performed better, with the model predicted R2 improved by 0.028 and 0.018, respectively, and the RMSE reduced by 0.150 and 0.107 g·kg-1, respectively. Overall, the 1D-CNN1 model based on CARS performs the best, with the predicted R2=0.846 and the RMSE decreased by 0.150 g·kg-1, respectively 0.846, and RMSE=3.145 g·kg-1, which reduces the network load while improving the model accuracy, and also proves that small-size convolution outperforms a larger number of large-size convolutions for better acquisition of data features. The SOM content prediction model is established by CARS screening feature wavelengths combined with shallow CNN, which provides a method and reference for establishing a high-precision SOM content prediction model.

Accurately grasping the total nitrogen content of farmland soil is significant for evaluating soil fertility and applying nitrogen fertilizer reasonably. To comprehensively utilize the advantages of each single prediction Model, improve the overall prediction performance, reduce the variance of the model, and improve the robustness, this study takes farmland brown soil as the research object, and based on near-infrared and visible hyperspectral data, puts forward a Combined prediction model based on standard deviation. CPM was used to predict soil total nitrogen content. Savitzky-Golay smoothing and first-order differential transformation are applied to the original hyperspectral data, and a tree model is used for feature band extraction. Using five single prediction models, Decision Tree Regression (DTR) (Model 1), Gaussian Kernel Regression (GKR) (Model 2), Random Forest Regression (RF) (Model 3), LASSO Regression (Model 4), and Multi-Layer Perceptron (MLP) (Model 5), a combination prediction model is established through a linear combination of single prediction models. The results indicate that: (1) The weights of the five single prediction models in the combined prediction model are obtained by generalized reduced gradient optimization algorithm: ω*1=0.407，ω*2=0.378，ω*3=0.215，ω*4=0，ω*5=0; (2) For all data, the predictive effectiveness of five single prediction models and combined prediction models for predicting soil total nitrogen content is M, respectively M1=0.855，M2=0.856，M3=0.847，M4=0.785，M5=0.796，MCPM=0.880, compared to the maximum predictive validity of a single model, the predictive validity of the combination prediction model has increased by 2.924%; (3) For all data, the prediction accuracy and standard deviation of soil total nitrogen content based on five single prediction models and combined prediction models are E(A1)=0.924，σ(A1)=0.075，E(A2)=0.928，σ(A2)=0.077，E(A3)=0.923，σ(A3)=0.082，E(A4)=0.882，σ(A4)=0.109，E(A5)=0.889，σ(A5)=0.104，E(ACPM)=0.937，σ(ACPM)=0.066, compared to the maximum prediction accuracy of a single model, the combination prediction model has improved prediction accuracy by 0.970% and model stability by 12.000%, making it an optimal combination prediction model. The combined prediction model can effectively estimate the total nitrogen content of farmland brown soil based on visible-near-infrared spectral data and can provide a basis and reference for the rapid monitoring of the total nitrogen content of farmland soil.

China has a huge and intensive open-field vegetable production system. However, serious issues such as low water and nitrogen use efficiencies, as well as excessive fertilizer application, limit the sustainability of the system. To improve the efficiency of production and enhance accurate fertilization in large-scale vegetable cultivation systems, this study was conducted with open-field lettuce. Three treatments of no nitrogen (N0), low nitrogen (N1) and high nitrogen (N2) were established. An unmanned aerial vehicle (UAV) equipped with a multi-spectral camera was used to establish correlations between three multi-spectral vegetation indices (NDVI, RVI, and SAVI) and lettuce chlorophyll content, biomass, crop nitrogen uptake, and total nitrogen content. Models to predict total nitrogen content for single growth stage and multiple growth stages were developed. The results showed that: (1) during rosette and heading stages, NDVI, RVI and SAVI values increased with the amount of applied nitrogen, but that during harvest stage, maximum values occurred with the N1 treatment; (2) NDVI showed a significant correlation with lettuce yield, nitrogen uptake and chlorophyll content during heading stage; and the total nitrogen content of lettuce was significantly correlated with chlorophyll content at p<0.01 level, with a correlation coefficient (R) of 0.51. When considering multiple growth stages together, NDVI values showed a significant correlation with lettuce yield, chlorophyll content, nitrogen uptake, and total nitrogen content at p<0.001 level, with correlation coefficients of 0.85, 0.82, 0.81, and 0.71, respectively. (3) Relationships of exponential, linear, logarithmic and power functions were fitted to the corresponding datasets, and the best prediction model of total nitrogen content for lettuce (N%=16.52ln(NDVI)+73.514) was established in the multiple growth stages. Using the lettuce total nitrogen content prediction model to obtain modeled values of total nitrogen content for the area of a commercial production field on the same farm, the average relative error was 3.2%, RMSE=0.556 6, NRMSE=0.010 8, showing accurate estimation of total nitrogen content. The results show that the model had good accuracy and that it is feasible to diagnose vegetable nitrogen content using unmanned aerial vehicle multi-spectral remote sensing.

The fruiting body of edible fungi is sensitive to light in the development stage, and there are no studies on the response of growth and nutritional quality of Pholiota adiposa to different spectral irradiation. In this study, Pholiota adipose was grown in an artificial light-type plant factory with pure white light as control (CK), and four treatments were set up: pure green (G), pure blue (B), pure red (R), and red-blue (RB). The light cycle was 12 h light /12 h darkness. The optimum light quality was determined by analyzing the effects of different light qualities on the traits, spectral characteristics, and mineral elements of Pholiota adiposa. The results showed that red light significantlypromoted the growth of Pholiota adiposestipe and the increase of fruiting body quality by 78.4% and 90.0%, respectively, compared with the control (p<0.05). Blue light significantly increased stipe diameter and pileus diameter by 22.8% and 19.1%, respectively, compared with the control (p<0.05). There was no significant difference in the thickness of the pileus between different treatments, indicating that the light quality had little effect on the pileus thickness. Compared with the control, the Hue value of Pholiota adiposepileus was increased in all treatments (5.3%~28.9% increase), while the Hue value of the stipe was decreased in all treatments (26.3%~46.7% decrease). The color spectral parameters such as C, MACRI, and PRI value of the pileus and stipe of Pholiota adiposa increased under green light treatment, which is more conducive to the coloring of Pholiota adiposa than other light qualities. In actual production, the shapes and colorationof the stipe and pileus can be adjusted by changing the light quality according to different market demands. Compared with the control, all light treatments increased the content of P and K-elements in the Pholiota adipose pileus (by 4%~15% and 7%~16%, respectively) and decreased the content of K, Ca, Mg, Na, and Mn elements in the stipe to different degrees. All light treatments decreased the accumulation of Ca and Na elements in the fruit body of Pholiota adiposa, and the accumulation of other elements except Ca and Na was increased and reached the highest level under green light. Therefore, green light is more favorable to mineral element accumulation in Pholiota adiposa than other light qualities. This study provides a theoretical basis for regulating the light environment in the factory production of the specialty mushroom Pholiota adiposa.

The anisotropic detection technology of crystals is gradually developing towards non-contact and non-destructive methods. Terahertz radiation has broad prospects in studying birefringence of anisotropic materials due to its large penetration depth and non-ionization characteristics for many dielectric materials. Quartz, sapphire, liquid crystal, and metamaterials containing sub-wavelength structures exhibit terahertz birefringence. As a common material in polarization functional devices, the parameter measurement is of great significance for developing terahertz devices. The extraction of material birefringence often depends on previous knowledge, such as optical axis direction and crystal thickness. The materials optical axis direction characterizes its anisotropys preferred direction. The appropriate Jones vector can be selected according to experience for crystals with known optical axis orientation. In practical applications, only the terahertz waves linear polarization direction, optical axis, and detection axis can be selected. It is easy to measure the ordinary and extraordinary light and calculate their refractive index directly from the time domain signal. For materials with unknown optical axis direction, it is necessary to rotate the sample to measure in different orientations. In addition, the extraction of birefringence depends on the thickness of the materials. The measured value obtained by vernier caliper or micrometer is quite different from the true value, and it is easy to cause scratches on the samples surface. At the same time, whether it is sample rotating or thickness measurement, human operation will introduce uncertainty for birefringence characterization. Based on the Terahertz time-domain spectroscopy (THz-TDS), a non-contact measurement method for the optical axis direction and the thickness of the birefringent crystal is developed in this paper. The complete refractive index properties of the crystal can be obtained without relying on the crystals prior parameters.The automatic positioning of the optical axis is realized by controlling the rotation of the sample and the action of the optical delay line. The iterative approximation algorithm of the transfer function is used to extract the crystal thickness and complete refractive index information.To validate our method, the (10-10) oriented sapphire, which exhibits birefringence at terahertz frequencies, was selected. We extracted the average extraordinary and ordinary refractive indices of sapphire in the frequency range of 0.3～1.5 THz, which is 3.08±0.02 and 3.39±0.02, respectively. The birefringence is -0.31±0.02, and the absorption spectrum was plotted. The results show that our method avoids sample damage and positioning errorscaused by manual operation and improves the efficiency, stability, and accuracy of terahertz birefringence extraction. It is significant to the polarization-sensitive terahertz measurement technology and its application.

The plasma generation and evolution processes are susceptible to matrix effects, environmental noise, pulse energy jitter, etc., resulting in the instability of the spectral data, which makes it difficult to ensure the validity of the monitoring criterion established based on a single spectral line. The monitoring criterion established by continuous multiple LIBS spectra combined with statistical methods can effectively improve the monitoring accuracy of paint removal based on LIBS technology. Based on a high-frequency nanosecond infrared pulsed laser paint removal LIBS online monitoring platform, the paper collected continuous multiple LIBS spectra of the paint removal process in real-time. After the spectral data were pre-processed by baseline correction and normalization, the spectral peaks of Ba Ⅰ (712.55 nm), Cr Ⅰ (357.48 nm), Cr Ⅰ (425.43 nm), Ti Ⅰ (427.45 nm), Cu Ⅱ (309.76 nm), Cu Ⅰ (484.22 nm) were used as the characteristic spectral lines and the paint removal effect was monitored. The intensity variation of the six characteristic spectral lines in different paint removal effects was studied, and the mapping relationship between different paint removal effects and the intensity variation of the selected characteristic spectral lines was established. The intensity of the above 6 spectral lines for each spectrum is extracted as data units. The data units of 10 consecutive spectral lines are used as data sets, and the data sets of each 10 iterations of the paint removal process are called data flow disks. The data cells and data sets in the data flow disk are analyzed. The confidence intervals are combined to determine the paint removal effect in the paint removal area in real-time. The monitoring criterion based on the data flow disk is obtained. The results show that this criterion can effectively monitor the paint removal effect in five categories: still on the top coat, completely removed top coat, still on the bottom coat, completely removed bottom coat, and substrate damage. The three-dimensional micro-pattern analysis showed that the accuracy of complete topcoat removal reached 1.2 μm, which effectively verified the applicability and stability of the LIBS-based data flow disc monitoring criterion.

Dissolved organic matter (DOM) is one of the important factors affecting the ecological environment and the safety of residents lives. When the total amount of DOM reaches a certain level, it will lead to the explosive growth of algae through water eutrophication, which makes the DOM composition more complex and the impact more severe. Although the standard detection methods can qualitatively analyze DOM, there are always bottlenecks in determining DOM components, and it is difficult for a single sensor to complete the complicated test of the total amount and components of DOM in water. Therefore, a DOM test scheme is proposed based on the cross-sensitivity between SPR (Surface Plasmon Resonance, SPR) sensors. Three classifiers are constructed using the BP (Back Propagation, BP) neural network trained by the swarm intelligence algorithm (Particle Swarm Optimization, PSO). The multi-mode fibers are coated with seven kinds of gold films with different thicknesses of 55~85 nm to form the SPR sensing probes with different optimal refractive index measurements to ensure the best refractive index measurement value of each sensing probe is effectively distributed in the range of 1.33~1.43 RIU, and each sensing probe has good sensitivity and linearity in the best measurement range. In the measurement range corresponding to other sensing probes, there are cross-responses as sensitive as possible through the change of wavelength, spectral width and light intensity. Finally, combined with the intelligent algorithm based on three classifiers of PSO-BP, through the experimental steps of water sample preparation of DOM, determination of DOM composition, measurement of refractive index and SPR effect, training and verification of artificial intelligence network, to realize the comprehensive training of the resonance wavelength, spectral width and light intensity of SPR effect on the measured samples. Thus, five DOM components (tyrosine proteins, tryptophan proteins, fulvic acid, soluble microbial metabolites, humic acids) of the inner canal(A), Hongze Lake (B), Park Landscape Lake (C) and campus landscape Lake (D) and their concentrations are effectively tested. Among them, the highest prediction rate for the concentration of DOM components in four different water bodies (A, B, C, D) is 81.2% (tyrosine), 85% (Tryptophan), 82% (Tryptophan), and 82.6% (Tryptophan), respectively. At the same time, the influence of the response parameters and the number of classifiers on the prediction effect is investigated, and the results show that the three classifiers have the best prediction effect compared with the two classifiers and the single classifier. The prediction accuracy of five different DOM component concentrations in mixed water sample can reach 81.5%, 84%, 81%, 82%, and 68.3%, respectively, to verify the correctness and feasibility of the multi-classifiers based on PSO-BP and fiber SPR sensor in DOM components prediction.

The analysis of ancient painted pigments is an important content of technological archaeology and cultural relics conservation research, which has important academic value and practical significance for exploring the development of ancient pigment technology and scientific conservation of cultural murals. Most of the traditional pigment identification algorithms are aimed at the pure pigments on the surface of painted cultural murals, whose identification accuracy is relatively poor for the mixed pigments on the surface of cultural relics. At the same time, chemical analysis methods usually require sampling of the surface of murals, which can easily cause damage. Hyperspectral technology is an emerging technology that has developed rapidly in recent years and has wide application in material identification. A method of mixed pigment identification based on hyperspectral intervals is proposed. Firstly, the first-order derivative of the reflectance spectrum of the unknown pigment is calculated, and the characteristic subinterval range is determined according to the “bump” of the first-order derivative curve of the unknown pigment. If the number of subintervals exceeds 2, only the two most obvious “raised” subintervals will be retained. Secondly, the reflectance curves of the unknown pigment and the standard pigment are transformed to the absorption-scattering ratio (K/S) using the KM model, which aligns more with the linear mixing characteristics. These K/S curves are normalized to［0, 1］ within the characteristic subinterval range. We calculate the similarity between the K/S curves of the unknown pigment and the standard pigments using the Spectral Angle Cosine combined with the Normalized Euclidean Distance. The top three results with the highest similarity for each characteristic subinterval are selected. Finally, one standard pigment K/S curve is removed from the identification results for each characteristic subinterval. Different standard pigment K/S curves from different characteristic subintervals are combined individually to generate the collection of subinterval identification results. They are combined with the abundance matrix obtained from the Dirichlet distribution function to generate 1 000 simulated mixed K/S curves. The similarity between the simulated mixed K/S spectra and the unknown pigment K/S curves is calculated again. We select one standard pigment with the highest similarity value in each collection. The similarity is compared again, and only the pigment with the highest value is identified as the final pigment for the unknown pigment spectra. The pure pigment samples were made by selecting Azurite, Malachite, Orpiment, and Cinnabar pigments. Six sets of mixed pigment samples were made by mixing pure pigments one by one. After the samples were drawn, the imaging data was collected by a hyperspectral imager. The feature subintervals are extracted, and the pigments are recognized by the proposed method after pre-processing. All the results were identified correctly except the Cinnabar in the mixed samples of Malachite and Cinnabar. The overall recognition rate for the mixed pigment samples is 83.3%. The results show that this method can identify mixed pigments and has practical significance for analyzing cultural relics pigments.

The expansion of Phyllostachys edulis (Moso Bamboo) into Cunninghamia lanceolata (Chinese fir) forest has led to ecological and economic problems such as “forest retreat and bamboo advance” and forest rights disputes. The use of remote sensing to effectively invert the succession process of Moso bamboo is of great significance for scientific control of forest resources. To reveal the effectiveness of canopy hyperspectral inversion of leaf area index (LAI) at different expansion stages of Moso bamboo to Chinese fir forest, four types of sample squares in mixed forest, which were divided based on a percentage of Moso bamboo were set to simulate expansion stages Ⅰ, Ⅱ, Ⅲ and Ⅳ along the expansion direction. Meanwhile, to explore the applicability of LAI hyperspectral inversion models for different expansion stages of Moso bamboo to Chinese fir forest, five types of single-factor regression models were established based on the characteristic wavebands which were chosen by the correlation between the original spectra, 10 spectral transformations such as open square, logarithmic, inverse, first-order differential and second-order differential and LAI at different expansion stages, and seven LAI significantly related vegetation indices such as normalized difference vegetation index (NDVI), yellowness index (YI) and so on. Multi-factor regression models were established based on the vegetation indices using four machine learning methods: neural network, decision forest regression, Bayesian linear regression, and linear regression. The results showed that the differential transform, including first-order differentiation (R′), second-order differentiation (R″), logarithmic first-order differentiation ［(lgR)′］ and inverse first-order differentiation ［(1/R)′］ of the original spectrum (R), could enrich the spectral information to characterize the expansion process better. Among the vegetation indices, YI had the highest correlation coefficient with LAI, showing high sensitivity to the expansion process, and NDVI had the best inversion effect. However, the overall inversion based on traditional vegetation index modeling was in effective. Moreover, the quadratic polynomial and power exponential regression models based on the differential transform spectra performed better in each expansion stage, while the inversion effect based on the vegetation indices was poor, and the neural network algorithm outperformed other machine learning algorithms. Traditional regression algorithms based on spectral transformations performed better than machine learning modeling methods. The model (y=5.291 4e183.76x) based on log-inverse first-order differential transform spectra fitted best in expansion stage Ⅲ with R2 of 0.735 and 0.742, RMSE of 0.733 and 0.468, and nRMSE of 14.0% and 9.9% for the modeling and validation sets, respectively. We suggest that the Moso bamboo expansion control should be selected in mixed forests of half of each species. Innovative analysis of LAI hyperspectral inversion modeling of Moso bamboo at different expansion stages will provide a basis for scientific silvicultural management.

Revealing the differences in foliar traits and leaf spectral characteristics between new and old leaves is important for the non-destructive monitoring of vegetation physiological and ecological parameters and can provide more theoretical support for quantitative remote sensing in forestry. This study sampled new and old leaves of 8 tree species in Changbai Mountain, and their reflectance spectra were measured. Then, the first-order derivative transformations and spectral indices were calculated. Foliar traits such as specific leaf area (SLA), leaf water content, leaf nitrogen content, and leaf carbon content were measured in our laboratory. Using variance analysis and correlation analysis, the differences in physicochemical properties and spectral characteristics between the old and new leaves of different tree species were investigated, and the differences in related coefficients were also analyzed between the old and new leaves. The results show that: (1) Multiple foliar traitsof thetree species showed significant differences between old and new leaves. Except for leaf carbon content, which did not differ significantly between old and new leaves, the other three traits showed significant variability between old and new leaves. (2) The differences in the spectral characteristics of different tree species were inconsistent between the old and new leaves. Only the old and new leaves of Betula costata, Ulmus laciniata, Acer buergerianum, and Pinus koraiensis showed more obvious differences in spectral curve characteristics. Betula costata showed significant differences in the spectral trilateral characteristics. (3) The correlation between leaf traits and spectra showed significant differences between old and new leaves, and the spectra have different abilities to indicate leaf traits. Near-infrared spectrum spectral reflectance is a better indicator of leaf nitrogen content for old leaves than for new leaves. In contrast, many spectral indices indicate better water and leaf carbon content in newer and older leaves. This study shows that there are not only differences in leaf properties and spectral characteristics but also differences in their correlations between old and new leaves. This study has a guiding significance for selecting representative leaves in the non-destructive observation of forest leaf properties.

Leaf pattern decoration on black glaze porcelain in High temperature,a cultural symbol combining elegance and aesthetic features, has gradually become a favored ceramic utensil in Chinese tea ceremony culture because of its unique decorative effect. However, its blending technology is still mysterious, and its formation and mechanism have not been fully disclosed under the continuous attention of scholars. Therefore, the mechanism of black glaze color and leaf pattern was revealed by using an experimental simulation method combined with spectroscopic analysis technology, and the relationship between glaze micro-structure and black glaze color was observed and analyzed by using sample hue angle h°, saturation C*, reflection spectrum curve and ultra-depth of field microscope. Based on traditional black glaze, a group of optimum glaze firing parameters was screened out, in which the Si/Al molar ratio was 10.14~10.47, MgO molar amount was 0.19~0.22, CaO molar amount was 0.42~0.47, and glaze thickness was 0.7~0.9 mm, which made black glaze show the most stable color and clarity, and leaf pattern decoration was ideal. The results show that the reasonable composition of glaze and glaze thickness is very important for the color presentation of black glaze and the clear decoration of the leaf pattern. At the same time, the spectroscopic analysis method and experimental simulation method adopted in this paper can effectively assist the research and development of glaze formula optimization in the ceramic industry, which is expected to improve the refinement and intelligence level of the porcelain-making process. At the same time, it realizes the analysis of microstructure and reflection spectrum curve characteristics of leaf pattern decoration on black glaze porcelain, highlights the application prospect of spectroscopic analysis method in the research and development of ceramic clay glaze, promotes the development of ceramic culture, and provides theoretical support for the protection, inheritance, and development of precious cultural heritage. It has important practical application value.

Apples are crisp, sweet, and inexpensive. They have great economic value in China and are a characteristic pillar industry for rural revitalization in apple-producing areas. Moldy core disease is one of the main internal defects of apples, greatly affecting their quality. Failure to remove it promptly will be detrimental to apple branding and seriously affect the development of the apple industry. In this study, near-infrared spectroscopy was used to obtain the spectra of apples in different poses and illumination modes. Pretreatment methods such as detrending, baseline correction, and second derivative (SD) were used to pre-process the original spectra, and then a support vector machine (SVM) was used to establish a moldy core disease discrimination model for apples. Also, the effect of the illumination modes on mold core disease detection in different poses was analyzed, and the apple mold core disease detection method, by fusing the spectral information of different illumination modes, was investigated. The results indicate that in the single-illumination mode, the 180°illumination mode has a higher correct discrimination rate for regular apples, while the 135°illumination mode has a better correct identification rate for moldy core disease apples; for the single-illumination mode, the optimal SVM model is obtained in the rightward stance of the fruit pedicel and the 135° illumination mode. The models sensitivity, specificity, and correctness in the calibration set are 1, 0.978 2, and 0.986 3, and the sensitivity, specificity, and correctness of the prediction set are 0.888 8, 0.956 5, and 0.937 5, respectively. For the information fusion of different illumination modes, most of the information fusion of the illumination mode can improve the performance of the apple mildew discrimination model to some extent; the optimal SVM model is obtained in the information fusion of 180° and 135°illumination modes in the rightward stance of the fruit pedicel, the sensitivity, specificity, and correctness of the model in calibration set are all 1. The prediction sets sensitivity, specificity, and correctness are 0.888 8, 1, and 0.968 7, respectively. Compared with the optimal model of single illumination mode, the models performance is improved. This paper provides a new idea of rapid, nondestructive identification of apple moldy core disease by near-infrared spectra with the information fusion of different illumination modes, and the method can also provide a reference for nondestructive detection of the internal quality of other fruits.

Water vapor is a key component of the atmospheric water cycle, influencing global cloud distribution and precipitation frequency. As precipitation significantly impacts climate and the environment, rapidly acquiring atmospheric water vapor concentration is critical in environmental climate research. The multi-axis differential optical absorption spectroscopy (MAX-DOAS) technique is a remote sensing method that enables fast and accurate measurement of trace gas concentrations in the atmosphere. Due to its stability, real-time online measurement, and multi-component and non-contact measurement advantages, this technique has become a promising new method for measuring atmospheric water vapor column concentration. Considering the narrow absorption band range of water vapor and the saturation absorption effect at high concentrations, this paper develops a MAX-DOAS water vapor vertical distribution detection system. It conducts an inversion algorithm study to retrieve the vertical column water vapor concentration accurately. The water vapor vertical column concentration was obtained at the study site in Huaibei. During the inversion process, the solar spectra collected in the zenith direction are chosen as the reference spectra. Using the Differential Optical Absorption Spectroscopy (DOAS) algorithm, the differential slant column densities (dSCD) of water vapor at different elevation angles are obtained. Finally, water vapors vertical column density (VCD) is extracted. The air mass factor (AMF) was obtained through geometric approximation. To minimize the interference from other gases, we analyzed the inversion errors for different spectral bands before the experiment and determined the optimal inversion band to be 433~452 nm. Continuous observations were conducted in the Huaibei region from February 24, 2023, to April 2, 2023. The experimental results indicate that during the monitoring period, the water vapor concentration in the Huabei region exhibited a V-shaped diurnal distribution pattern, with higher concentrations in the morning and evening and lower concentrations around noon. A correlation analysis was performed between water vapors observed vertical column density (VCD) and the daily reanalysis data from the European Centre for Medium-Range Weather Forecasts (ECMWF)ERA5. The results showed a good consistency between the two datasets (R=0.95). Furthermore, an analysis of the relationship between wind speed, wind direction, and H2O VCD distribution during the monitoring period revealed that when the wind direction was around 60° and the wind speed was less than 5 m·s-1, there was an increasing trend in H2O VCD. During the pollution stage, low wind speeds and higher water vapor concentrations were commonly observed features. The study demonstrated that the ground-based MAX-DOAS system effectively monitored water vapor vertical column concentration in the blue light band, providing an effective technical means for inverting atmospheric water vapor vertical column concentration.