The X-ray Photoelectron Spectroscopy (XPS) technique can provide information about the chemical states of various elements on the sample surface and the peak intensities and positions of their spectra. The thickness of thin films can be calculated by utilizing this information and the equation for photoelectron signals. This paper introduces three methods for handling the equation of photoelectron signals: direct solving method, substrate-ratio method, and angle-ratio method. Their derivation processes are analyzed, and the results indicate that these three methods have different accuracies and applicability ranges. The direct-solving method has the widest applicability but the lowest accuracy. The substrate-ratio method has the most limited applicability range, being able to calculate the thickness of thin films only on substrates of infinite thickness, and the effective attenuation length of the composition of film and substrate composition needs to be similar. However, it is least affected by instrumental and carbon contamination errors, thus having the highest calculation accuracy. The angle-ratio method has a moderate applicability range and calculation accuracy, and it can be used without considering the limitation of the substrate layer. However, the variation in the emission angle significantly affects the calculation accuracy. Researchers can consider the above factors comprehensively when calculating the thickness of thin films using XPS data.

The diagnosis and treatment of tumors have beena hot medical research area for a long time. Terahertz technology has great potential for application in the biomedical field due to its low photon energy, label-free, and high temporal resolution. This paper introduces the application of terahertz technology in tumor diagnosis and treatment. Terahertz tumor diagnosis is mainly based on terahertz spectroscopy and imaging technology. By using terahertz spectroscopy to obtain spectral feature information, with the combination of advanced pattern recognition and machine learning algorithms to extract high-dimensional features, tumors can be detected and diagnosed. The imaging technique mostly utilizes the differences in components between tumor cells and tissues to distinguish and identify tumors. The research on terahertz technology in tumor diagnosis is first introduced in three dimensions: biomolecules, cells, and tissues. The research on biomolecules focuses on tumor markers such as proteins, nucleic acids, amino acids, and glycans. Advanced research in cellular detection focuses on identifying tumor cells and detecting blood cells. Regarding tissue, the detection and diagnosis of cancer tissues with different types is presented in two main dimensions: imaging and terahertz time-domain spectroscopy. The research on tumor therapy based on terahertz technology emphasizes the positive effects, including biological effects on ablation of tumor cells and modulation of gene expression to improve body functions. Finally, the current limitations and prospects of terahertz technology in biomedical applications are discussed to provide insights for future research on terahertz in related fields.

This study determined the average Raman shifts (Ag/cm-1) of pyrites in three stages, Py1, Py2, and Py3, in the Woxi gold deposit, which were 378.3, 372.5, and 380.3, respectively, indicating the deposit have gone through a process of high pressure, low pressure，and high pressure; The average full width at half maximum (FWHM/cm-1) of the characteristic Raman peaks for Py1, Py2, and Py3 were 8.6, 10.8, and 9.1, suggesting that thetemperatureof ore fluids underwent a process of low temperature, high temperature, and low temperature. Based on previous research and regional tectonic evolution history, it inferred that Py1 was formed under the compressional background of the Late Jurassic, with ore fluids sourcing from regional metamorphic; Py2 was formed under extensional background of the Early Cretaceous, with ore fluids originating from post-magmatic hydrothermal activity which associated with large-scale magmatic events in extensional environment; Py3 was formed under the compressional background of the Early- to Middle Eocene, with ore fluids sourcing from regional metamorphic. By integrating the Raman characteristics of pyrites with the age of the main mineralization stage and the regional tectonic evolution history, it is possible to constrain further the ages of pyrites formed at the other stages, explore the formation process of deposit, extract ore-forming information, and expand the application of laser Raman spectroscopy technology in the field of geology.

To improve the detection effect of laser-induced breakdown spectroscopy on the metal elements in lubricating oil and avoid the problems of plasma quenching, oil splashing, and low spectral intensity when laser-induced breakdown spectroscopy is applied to the detection of lubricating oil, beeswax was used as the matrix to convert the lubricating oil sample from the liquid phase to the solid phase. The Mg element and Ca element in the lubricating oil were quantitatively analyzed. Firstly, a scheme was proposed to prepare lubricating oil test samples using beeswax. Secondly, the experimental parameters such as pulsed laser energy, acquisition delay, and laser focus position were optimized. The laser energy was adjusted from 30 to 120 mJ. Each time, it was increased by 10 mJ, and the effect of laser energy on the spectral line intensity and signal-to-back ratio was compared and analyzed. The results showed that the experimental results were the best when the laser energy was 90 mJ. The best experimental results can be obtained with an acquisition delay of 2.5 s, choosing different acquisition delays, varying by 0.5 s each time, and comparing the experimental results with acquisition delays ranging from 1 to 5 s. The influence of the laser focus position on the spectral signal was compared and analyzed, and the laser focus position was adjusted from 0.5 mm above the sample surface to 5 mm below the sample surface. By moving 1 mm each time, it was concluded that when the laser focus position was 2 mm below the sample surface, the target element’s spectral signal intensity and signal-to-background ratio were the best. Then, Mg(Ⅱ) 279.552 8 nm and Ca(Ⅱ) 393.366 nm were selected as the analytical lines of Mg and Ca. Under the best experimental conditions, seven lubricating oil samples prepared with different concentrations of beeswax were spectra collected, and the calibration curves of Mg and Ca were established. The linear correlation coefficient of the calibration curve of Mg and Ca reached 0.996 1 and 0.995 8, and the detection limits of Mg and Ca were 4.08 and 6.11 g·g-1. Finally, based on the established calibration curves, the concentrations of Mg and Ca in four other lubricant samples with different concentrations were detected, and the recoveries of Mg and Ca were 92.67%~106.15%. The recoveries of Ca were 95.88%~108.57%. The results show that the use of beeswax as a matrix to prepare samples for the detection of metal elements in lubricating oil solves the problems of low spectral intensity and oil splashing when laser-induced breakdown spectroscopy is used for lubricating oil detection and realizes the detection of metal elements in lubricating oil on the order of g·g-1.The proposed method is of great scientific significance for detecting metal elements in lubricating oil by laser-induced breakdown spectroscopy.

Raman spectra are scattering spectra based on the Raman scattering effect. Since the vibration and rotation energy characteristics of different kinds of substances are unique, the resulting Raman scattering spectra are also unique. Raman spectroscopy is very advantageous in identifying the composition of substances. It is also favored for its lossless, non-contact, fast, simple, and repeatable characteristics and is widely used in various fields such as chemistry, physics, biology, and medicine. However, due to the weak signals measured, the processing accuracy of optical instruments, and the interaction between the components of the mixture, the Raman spectra of the mixture not only have the phenomenon of overlapping peaks but also some of the characteristic peaks of the weaker signals may be submerged in the background noise, which affects the accuracy of Raman spectroscopy analysis of mixtures. This study applies principal component analysis to Raman spectral analysis to solve the difficulty of analyzing and identifying the weak signals in Raman spectra. It proposes a Raman signal extraction method based on full spectral information. In this method, the measured Raman spectra are regarded as the linear superposition of the spectra of different material components, and the Raman signals of different material components are extracted through the principal component analysis of multiple Raman spectra with different component ratios, separating the background noise and random noise. According to the characteristics of Raman spectra of material components, which are not necessary to satisfy orthogonality, this paper analyzes and discusses the relationship between spectral principal components and Raman spectral components of material components and gives a general method of using the spectral principal components to be corrected to Raman spectra of material components. In addition, according to the linear correlation characteristics between the spectral principal components and the concentration of the material components, this paper also gives the basis for determining the Raman spectra of the material components, the linearity error, and the random noise. Through the experimental verification of Raman spectra of methanol and ethanol mixed solutions with different concentrations, the extraction of methanol and ethanol Raman signals is realized. The background noise and random noise are successfully separated. The final results match the reference signal well. The judgment results of Raman signals of the material components, linearity error, and random noise are verified simultaneously. In this paper, an effective method of extracting actual spectral components using Raman spectral principal components is proposed, which has the advantages of being fast and convenient, low cost and high accuracy, and is a useful supplement and attempt to the Raman spectral data processing technology, and has great potential for application in substance identification and concentration detection.

Spatial Heterodyne Spectroscopy（SHS）is a new hyperspectral remote sensing detection technology widely used in atmospheric observation, astronomical remote sensing, material identification, and other fields. Two-dimensional measured interferometric data acquired by SHS can be interfered with by various influences, of which high-frequency noise, irregular dark spots, and interferogram nonuniformity are among the most common. These effects reduce the accuracy of the recovered spectra, and therefore, effective data correction methods need to be developed for these effects to improve the accuracy of the inverted spectra. In this paper, two light sources, potassium and xenon lamps, are used to generate quasi-monochromatic and continuous light signals, and the interference data formed by them are used as the object of study. A spatial heterodyne interferogram data correction method based on principal component analysis is proposed to address the effects of multiple noises in these two measured interferograms. Firstly, the first-order difference method is used to preprocess all the row data of the measured interferograms to remove the baseline effects, and Fourier transforms the processed row data to obtain the spectral data. Then, all the line spectral data are subjected to principal component analysis, multiple mutually orthogonal principal components and the contribution of each principal component is calculated, and the principal components with a contribution of less than 2% are treated as noise and deducted. In contrast, the other principal components are retained as valid spectral signals for spectral reconstruction, and the reconstructed spectra are inverse Fourier transformed to obtain a corrected interferogram. Finally, the effectiveness of the calibration methods is comparatively analyzed in terms of interferogram and spectral dimensions. The results show that the dark spots in the measured interferograms of monochromatic and continuous two light sources are effectively deducted, and the effect of non-uniformity is greatly improved. The effects before and after spectral correction are compared for the data in rows 536, 600, and 982 of the interferogram, which are affected by the dark spots. The results show that the correction method effectively suppresses the high-frequency noise in the spectra and makes the spectra smooth and clear, and the details of the characteristic peaks and so on are highlighted. The signal-to-noise ratio is improved, and the mean square error of the three rows of spectra decreased from 0.037 77, 0.027 33, and 0.030 99 before correction to 0.013 31, 0.012 20, and 0.012 34 after correction, respectively, which quantitatively illustrates the effectiveness of the method.

Laser-induced breakdown spectroscopy (LIBS) technology, as an effective tool for material composition analysis, has broad application value. However, due to the poor repeatability of LIBS and the influence of matrix and self-absorption effects, spectral data contains a large number of redundant features that are useless for quantitative analysis. To overcome the difficulty in improving prediction accuracy when using raw full spectrum data as model input, two feature engineering techniques (minimum absolute shrinkage and selection operator regression LASSO and sequential backward selection SBS) were combined with machine learning to achieve a quantitative analysis of nickel (Ni), titanium (Ti), and chromium (Cr) in stainless steel samples. This study used seven stainless steel samples with different element contents purchased from Steel Research Nanogram Testing Technology Co., Ltd. as the research objects. Seventy LIBS spectra were obtained, and four different data preprocessing methods were compared, including Maximum Minimum Normalization (MMN), Standard Normal Variation (SNV), Savitzky Golay Smooth Filtering (SG), and Internal Standard Method (IS). The preprocessing results were detected using Root Mean Square Error (RSME). Finally, Savitzky Golay smoothing filtering was chosen for spectral preprocessing. Effective variables were independently selected for different quantization elements when selecting features using LASSO and SBS algorithms. Then, three different feature combinations, namely full spectrum, LASSO selection feature, and SBS selection feature were used as inputs to the model. To verify the effectiveness of the feature selection method, partial least squares (PLS) Compare two different machine learning models using a Support Vector Machine (SVM). Evaluate the performance of different models using Average Relative Error (ARE) and Relative Standard Deviation (RSD). The results showed that the model inputs selected by the two feature selection methods showed better prediction accuracy and stability compared to full-spectrum inputs in different machine learning models. Among them, the LASSO-PLS model achieved the best prediction accuracy in the quantitative analysis of Ni, Ti, and Cr elements, with ARE of 3.50%, 2.66%, and 0.93%, and RSD of 4.55%, 5.23%, and 2.04%, respectively. Therefore, the LIBS combined with LASSO and SBS algorithms proposed in this article can accurately and stably quantify the Ni, Ti, and Cr elements in stainless steel, providing a reference for further exploring the application of LIBS combined with machine learning in stainless steel element quantification analysis scenarios.

Moisture content significantly impacts the properties (e.g., stability and compressibility) of chemical and pharmaceutical granular products. The traditional fluidized bed drying process moisture detection uses traditional instrumentation to detect the process of humidity, temperature, and other characterization variables and then infer the moisture content; this method often produces inaccurate detection, has a lag and other shortcomings, it has been difficult to meet the needs of modern production. Near-infrared (NIR) spectroscopy, as a new sensor technology, can be obtained from the molecular level of process information; its operation is simple, has fast analysis speed, and there is no need for sample pre-processing and other advantages, so it is widely used in many fields. However, existing NIR spectroscopic analysis methods are mainly based on offline detection of collected samples, which makes it difficult to reflect the real-time status of the production process. At the same time, in most cases, the absorption peaks of the collected NIR spectra overlap severely, resulting in the effective information of the NIR spectra being masked by various noises. Therefore, it is necessary to use suitable analysis tools for NIR data analysis and effective information extraction. Traditional algorithmic models mostly use linear or single-model methods, which makes it difficult to effectively solve the problem of effective information extraction from NIR spectra. Thus, in this paper, the fluidized bed drying (FBD) process of batch particles is used as the detection object, and near-infrared spectroscopy is applied to the fluidized bed granulation and drying process, which is combined with the XGBoost algorithm to establish an on-line measurement model of moisture content of particles. The Beluga whale optimization obtained the optimal parameters of the model, and then the validity of this approach was verified by the real fluidized bed drying experiments. For the validation experiments, the wave numbers (4 798 to 9 423 cm-1), which include the characteristic peaks of moisture and have more stable signals, are selected for modelling. Three independent batches of data out of the four batches collected are used as training sets to train the model, and the fourth batch is used to test the model. The models are evaluated in terms of Root Mean Squared Error (RMSE) and Coefficient of Determination R2 (R-Square), which show that the optimized XGBoost model outperforms the models built by PLS and BP-ANN algorithms in all the metrics. The online moisture content detection model based on near-infrared spectroscopy and XGBoost proposed in this paper provides a new approach for online moisture content detection in the fluidized bed drying process.

The magnesium borate system’s dynamic behavior was investigated using ATR-FTIR and a self-made continuous humidity adjustable device. The supersaturation transformation and crystallization process of the three kinds of Ascharite, including Hungchaoite, Mcallisterite, and Inderite, were studied. Continuous adjustment of the relative humidity of the environment in the pool was achieved, allowing continuous monitoring of the dynamic chemical behavior of the magnesium borate supersaturation during evaporation. The supersaturation liquid film prepared on the substrate surface gradually evaporated and condensed into an ultra-concentrated state, eventually forming a multiphase mixture containing various magnesium borate crystals. Continuous measurement of infrared spectra of magnesium borate supersaturation on the substrate surface and magnesium borate crystals obtained by phase transformation and crystallization, combined with density functional theory calculations of infrared vibration patterns of various magnesium borate crystals, the supersaturation of poly borate in the three magnesium borates, the distribution of poly borate ions during the phase transformation, and the types of magnesium borate crystals formed by the final crystallization were described, the difference of species distribution of borates in three kinds of magnesium borate systems was summarized, and the reaction mechanism of the whole crystallization process was given. Research results: (1) In the three supersaturated solutions of Ascharite, the main polyborate ion is B3O3(OH)52-. After evaporation and crystallization of the three supersaturated solutions of Ascharite, Hungchaoite B4O5(OH)42- can be observed in the crystals, which is a stable species formed by the polymerization reaction of B3O3(OH)52-; (2) The asymmetric stretching vibration of the three-coordinated boron-oxygen bond as(B(3)—O) in the magnesium borate crystal is mainly inward stretching. In contrast, the borate ions in the solution mainly stretch outward. The difference in vibration direction can be used as a basis to distinguish the presence of borate ions in crystals and solutions; (3) There is no B5O6(OH)4- in the magnesium borate salt obtained by the “dilution to salt” process, and in this study, B5O6(OH)4- only appears during the phase transition crystallization process of the supersaturated solution of Hungchaoite, which is explained by the fact that B4O5(OH)42- in the supersaturated solution of Hungchaoite is more likely to undergo polymerization reaction to generate B5O6(OH)4-. This study expands the traditional concentration range of borate system research. It describes the dynamic evolution process of magnesium borate in detail, which provides a new understanding of the infrared spectroscopy study of magnesium borate systems during phase change.

In photoacoustic spectroscopy, periodic modulation of light sources is one of the necessary conditions for generating photoacoustic signals. The modulation can only be realized using mechanical choppers for some light sources, such as infrared radiation sources applied in broadband photoacoustic spectroscopy. However, a commercial chopper is unsuitable for miniaturized photoacoustic spectroscopy instruments due to its large size and high cost. In this paper, we carried out the design and development of a homemade chopper and its application to carbon dioxide (CO2) detection by photoacoustic spectroscopy. The chopper control circuit was designed based on the STM32 microcontroller technology, and a comparison experiment between the homemade chopper and the commercial chopper was conducted. The experimental results showed that the photoacoustic signals produced by both had a good consistency, verifying the feasibility and reliability of the homemade chopper, which can satisfy the application requirements of photoacoustic spectroscopy. Then, the research of photoacoustic spectroscopy for measuring CO2 was carried out based on the homemade chopper, and the effect of humidity on the CO2 photoacoustic signal was analyzed. It has been shown that water vapor significantly accelerates the molecular relaxation rate of CO2, thus improving the amplitude of the photoacoustic signal of CO2. The Allan deviation result showed that the detection sensitivity for CO2 of the photoacoustic spectroscopy system was found to be 5 L·L-1 at an average time of 200 s. Compared with the CO2 before humidification, the minimum detection limit of CO2 was improved by a factor of 2.4 after humidification. The homemade chopper is characterized by easy integration and has reference or application value for developing high-sensitivity photoacoustic spectroscopy instruments.

Terahertz waves are electromagnetic waves with frequencies between 0.1 and 10 THz, between microwave and infrared, with strong penetration, high resolution, no ionizing radiation, and other advantages, which have a wide range of application prospects in the fields of safety detection, medical diagnosis, material analysis and so on. The demand for Terahertz technology development and application of terahertz wave multifunctional modulation devices is growing. The traditional medium in the terahertz band response is weak; the metasurface, because of its ability to modulate the terahertz wave at the wavelength scale and in the processing and design of low-cost, small volume and other advantages by the researchers are widely concerned. Terahertz functional devices with metasurface have been reported to have more or less single functions and are limited by a single polarisation wave. Researchers propose that the program is mainly to change the geometric morphology and arrangement of the sub-wavelength metasurface structure to introduce the transmission phase or the geometric phase through the rotating unit structure to achieve the electromagnetic wave parameter tuning. However, a single tuning method still exists due to poor tunability and so on, and therefore, the design of the metasurface electric field tuning device based on the synergistic tuning of the transmission and geometric phase theory seems more prospective and meaningful. In this paper, a new diagonal double-cross structure metasurface is introduced, which consists of a top metal pattern, an intermediate dielectric layer, and a bottom metal plate and combines the synergistic effect of the transmission phase and geometrical phase to complete the independent tuning of the left circularly polarised wave and the right circularly polarised wave. At the incident of left/right circularly polarised waves with a frequency of 1.1 THz, the metasurface exhibits a variety of functions, including vortex beams, beam splitting, vortex wave splitting, and focusing beams with different topological charges. This innovative structural design provides a new idea for researching multi-functional and multi-polarisation terahertz modulation devices. It offers potential application scenarios in the field of terahertz wireless communication. Future research can extend it to the microwave and optical fields by changing the metasurface dimensions.

Wood species can be classified according to thedifferences in spectral reflectance of different wood surfaces. To prevent wood corruption or cracking, glorify the appearance, and prolong the service time of wood products, wood surfaces often need to be coated with paints in the production practice of wooden furniture and handicrafts. The influence of paints on the spectral reflectance of wood surfaces will lead to the drift and transformation of spectral curves. The experimental results show that the classification model trained by the spectral reflectance of the original wood samples cannot be used to classify the painted wood. The drift and transformation of spectral curves of painted wood compared with original spectral curves without coated paints can be fitted by nonlinear models such as neural networks. To continue to use the classification model trained by the original wood dataset, we use a fully connected neural network to fit the relationship model between the original spectral reflectance and the coating spectral reflectance. Through this model and SVM classifier, the coating spectral reflectance is corrected to realize the classification of painted wood using the classification model trained by the original wood spectrum. We also use the convolutional neural network model to extract the convolutional features of spectral reflectance and add a hidden layer based on the relationship between the convolutional features of original spectral reflectance and coating spectral reflectance to modify the convolutional features of coating spectral reflectance and output the classification results through the output layer. To verify the effectiveness of the proposed scheme, we collect the near-infrared spectral reflectance (NIR: 950~1 650 nm) and visible/NIR spectral reflectance (VIS/NIR: 350~1 000 nm) of 20 wood species and compare the performances of the rectification models for 8 different paints in terms of the corrected spectra. Because of the experimental results, it can be concluded that the classification performance of NIR is better than that of VIS/NIR. The correction model based on the convolutional neural network can improve the NIR classification accuracy of the transparent wood surface to more than 70%. In contrast, the fully connected neural network can improve that to more than 80%, but neither model can correct non-transparent coatings on the wood surface. Regarding training speed and recognition efficiency, the correction model based on a convolutional neural network is better than that based on a fully connected neural network. In summary, the nonlinear relationship model between the original spectral reflectance and the coating spectral reflectance established by these two neural networks can correct the coating spectral reflectance with clear paints and then classify the painted wood directly by the classification model trained by the original spectral reflectance. The proposed scheme extends wood species classification from the original wood product to coating wood with clear paints with certain practical application significance and prospect.

4-nitrophenol (4-NP), a phenolic compound with highly toxic and carcinogenic properties, has become a global concern. As a fluorescence probe, a rapid and sensitive fluorescence assay for 4-NP was developed using cysteine-protected copper nanoclusters (Cys-CuNCs). The Cys-CuNCs were prepared using ascorbic acid-capped copper nanoparticles (CuNPs) as a precursor and cysteine as an etching agent. The prepared Cys-CuNCs showed excellent fluorescence properties and stability solubility with the maximum excitation and emission peak at 370 nm and 464 nm, respectively. In the presence of 4-NP, the fluorescence of CuNCs was quenched effectively. Single-factor experiments were investigated to optimize the pH of the buffer and the reaction time for nitrophenol detection. The optimal conditions were as follows: phosphate buffer (10 mmol·L-1, pH 8.0) as a working buffer, mixing 4-NP and CuNCs for fluorescence scan. A sensitive analysis method for nitrophenol was constructed by utilizing the variation of fluorescence intensity F0-F/F0 as abscissa and 4-NP concentration as ordinate. The fluorescence quenching rate of CuNCs showed a linear relationship with the concentration of 4-NP in the range from 1 to 100 mol·L-1 with the detection limit of 0.31 mol·L-1. Four other 4-NP homologues and eight other organics were selected as the interferent; all of them had almost no effect on the fluorescence of CuNCs, indicating that the method had good anti-interference ability. In addition, the strategy was successfully applied to detect 4-NP in lake water with satisfactory recoveries from 97.0%～101%. The relative standard deviation was 2.71%～3.60%, indicating that the method is accurate and could be used for 4-NP detection in actual samples.

At present, traditional liquor selection commonly employs the method of “liquor picking by flowers” during production, relying on workers’ subjective experience for evaluation. However, multiple influencing factors affect the actual production, resulting in uncertainty in the process of liquor connection, posing challenges in ensuring the stability of original liquor quality. This study collected samples of the original and composite original liquor with varying concentrations of tail liquor (0.0%~2.0%). The three-dimensional fluorescence spectra were obtained by fluorescence scanning, establishing a correlation between substance changes and fluorescence data. The fluorescence spectra underwent pre-processing steps such as removing scattering, Raman normalization, Savitzky-Golay smoothing, and removing outliers. Subsequently, parallel factor analysis was used to decompose the spectra into four uncorrelated components, and these components were initially identified through composite similarity analysis in conjunction with the attributes observed in single-substance fluorescence spectra. The results show a higher correlation between the fluorescence spectra of most acids and esters with component 2, suggesting that acids and esters have a stronger influence on the fluorescence properties of component 2. The dataset is reduced from 781×61×164 to 4×164, achieving data dimensionality reduction. A support vector machine (SVM) model was developed to assess the quality of the original liquor. A genetic algorithm (GA) was also employed to optimize the SVM model. GA-SVM model performs better than the original SVM model in accuracy and precision. The optimized model achieved an accuracy of 88.64% compared to 95.45% of the original model, and the precision improved from 0.94 to 1.00. This suggests that integrating 3-D fluorescence and chemometrics is an effective method for rapid detection to evaluate the quality of the original liquor. And provide support for online detection during the distillate liquor selection process, thereby enhancing the overall quality control.

Near-infrared spectroscopy has been widely used due to its high efficiency and non-destructive properties advantages. However, the consistency phenomenon between near-infrared spectrometers can lead to insufficient accuracy when the master model predicts the spectra of its slave instruments. If the calibration model is rebuilt based on the offset spectrum, it will lead to higher consumption. This paper proposes a transfer component analysis direct standardization (TCADS) algorithm to address the above issues.The algorithm initially employs an enhanced TCA algorithm to convert master and enslaved person spectra, which adhere to distinct distributions, by projecting them into high-dimensional reproducing kernel Hilbert space. Subsequently, it reduces the dimensionality of their spectral matrices. Finally, a direct standardization algorithm is reapplied to the master and slave spectra post-TCA transformation, further enhancing the model’s transfer performance. This algorithm combines nonlinear correction with linear correction, effectively alleviating the problem of overcorrection compared to traditional linear correction algorithms, and is robust. To verify the effectiveness of the algorithm, experiments were conducted on public datasets and compared with traditional direct standardization (DS), piecewise direct standardization (PDS), and slope and bias correction (SBC) methods. The experiment demonstrates that the TCADS algorithm proposed in this article efficiently minimizes spectral disparities between the master instrument and the slave instrument. This enhancement notably outperforms traditional model transfer algorithms, facilitating the effective sharing of near-infrared spectral models established on the master instrument to the slave instrument.

X80 high-grade pipeline steel is the main material used in long-distance oil and gas transmission pipelines. During the application process, it was found that there was a significant difference in the toughness of the weld metal after welding steel plates from different manufacturers, which seriously affected the safety of oil and gas storage and transportation. There was an urgent need for a fast and accurate in-situ quantitative analysis method for the distribution of elements along the thickness direction of the weld seam to help explore the mechanism of the joint effect of X80 pipeline steel base material and welding material on the toughness of the weld metal. Therefore, this article proposed a method for in-situ quantitative analysis of Mn, Ni, Cr, Al, and Nb in weld using LA-ICP-MS. By optimizing the laser pulse frequency to 20Hz, laser energy to 100% (laser output mode Image Aperture), etching aperture to 100m, and defocus distance 0m, the strength and the stability of mass spectrometry signals were enhanced. The experiment was calibrated using standard samples matched with the matrix, and the matrix element57Fe was used as the internal standard for correction. By analyzing related mass spectrometry interferences, isotopes27Al, 53Cr, 55Mn, 60Ni, and 93Nb were selected. The established LA-ICP-MS micro zone in-situ quantitative analysis method was applied to analyze the distribution of element content in two X80 pipeline steel welds with the same welding material but different base material compositions. The correlation coefficient of this method ranged from 0.992 7 to 0.999 6, with a quantification limit of 0.23~2.57 g·g-1. The results showed that Mn, Cr, Al, and Nb with similar contents in the two base metals exhibited similar dilution at the root of the weld. In comparison, Ni elements with significant differences in content between the two base metals showed significant differences in content within 8.4 mm from the root of the weld. The impact test results showed that the toughness of weld with high Ni element content in the base material is relatively significantly higher. SEM analysis of weld root showed that the increase of Ni element content was conducive to forming lath bainite structure. Therefore, it is considered that the dilution of base metal to Ni in the root weld metal can be reduced by adding 0.14% Ni to X80 pipeline steel. The weld impact toughness can be improved by higher Ni content by promoting the low-temperature lath bainite transformation. The established LA-ICP-MS in-situ quantitative analysis method is of reference significance for ensuring the safe operation of the X80 long-distance pipeline.

As a famous folk kiln in the northern central plains, the Hebi Ji kiln had a grand production scale, a wide variety of types, and exquisite craftsmanship, with white-glazed porcelain as its main production. Also, it produced a variety of glazed porcelains, such as black-glazed, Jun-glazed, yellow-glazed, and bean-glazed porcelains. In this study, the chemical compositions, microstructures, and reflectance spectra of the body glaze of Song Dynasty coarse white porcelain, incised flower coarse white porcelain, and Jin Dynasty coarse white porcelain from the Hebi Ji kiln were tested and analyzed using the techniques of proton-excited X-ray fluorescence (PIXE), polarized light microscopy (OM), spectrophotometer (UV-Vis-NIR) and thermal expansion meter (TD). The results of the study show that most of the three types of coarse white porcelain in the Hebi ji kiln show a faint yellow-green hue and are dominated by a yellow hue, with the reflectance of most of the samples distributed in the range of 43.4R%~57.2R%, and the glaze brightness value L* is generally high, distributed in the range of 72.3~90.3. Song and Jin dynasties’ coarse white porcelain glaze formula is relatively stable, of which the Jin Dynasty and Song Dynasty coarse white porcelain glaze formula is the same, carved flowers in the white porcelain glaze in addition to the total amount of flux relative to the other two types of low, the rest of the chemical composition is the same. Three types of coarse white porcelain glaze belong to the calcium alkali glaze system, and the glaze layer and the body between the existence of varying thicknesses of the cosmetic clay layer. Three coarse white porcelain carcass types have northern high alumina and low silicon characteristics; the Jin Dynasty coarse white porcelain carcass flux content is slightly lower than that of the two types of coarse white porcelain in the Song Dynasty, and the raw material characteristics are similar. During the Song Dynasty, the firing temperatures of coarse white porcelain were generally higher. Specifically, the firing temperatures for standard coarse white porcelain ranged from 1 250~1 340 ℃, while the carved coarse white porcelain was fired at temperatures between 1 310 and 1 390 ℃. Both types exhibited states of “the raw firing” and “the over firing” with generally good body density. In contrast, the firing temperatures for coarse white porcelain during the Jin Dynasty were lower, ranging from 1220~1250 ℃. These pieces were mostly in a "raw firing" state, resulting in poorer sintering. Hebi Ji kiln Jin dynasty coarse white porcelain and Song dynasty coarse white porcelain compared porcelain production process has declined. This study comprehensively analyses the process characteristics and development trends of the coarse white porcelain of the Hebi Ji kiln, which is of great value in understanding its relationship with other kilns and lays a scientific foundation for further future clarification of its relationship with other kilns.

In high-resolution astronomical spectral detection using multi-mode optical fibers for solar surface magnetic field and visual velocity measurement, there is a significant phenomenon of uneven energy distribution and speckle in the obtained spectral images. Analyzing the reasons, it can be concluded that during fiber optic spectroscopic imaging, the imaging system receives the image of the output fiber end at different wavelengths and spreads along the dispersion direction. The high dispersion rate of the spectrometer results in a very narrow spectral range, which has excellent coherence, corresponding to each pixel in the imaging system. There are multiple transmission modes in multimode optical fibers, and the energy center of the speckle pattern formed by interference between different modes at each wavelength will deviate from the fiber’s geometric center, reducing the accuracy of spectral measurement. To address this issue, this paper proposes a multi-dimensional mechanical perturbation system to suppress speckle effects and improve the accuracy of fiber optic spectral measurements. The multi-dimensional mechanical disturbance device consists of three mechanical structures with reciprocating motion in different directions and frequencies. By adjusting three mechanical devices at different frequencies, the conduction mode in the optical fiber undergoes a nearly random phase drift, resulting in a random change in the speckle pattern at the output end of the optical fiber. After long-term exposure (corresponding to superimposing and averaging multiple speckle patterns), the influence of speckles can be eliminated. To test the mode disturbance effect of the system, the energy distribution of the outgoing speckle field of a 650 nm laser was studied using a fiber with a core diameter of 35 m. The standard deviation of the energy center position was proposed as the evaluation function for the dispersion of the fiber mode speckle energy and its impact on the accuracy of astronomical fiber spectral measurement. Firstly, a comparative analysis was conducted on the average stacking effect of 1 000 speckle patterns under non-disturbance mode, manual disturbance mode, one-dimensional disturbance mode, and multi-dimensional disturbance mode. The results showed that manual disturbance mode and multi-dimensional mechanical disturbance mode had better effects, and the energy distribution of the average speckle pattern stacking was relatively uniform. Then, the multi-dimensional perturbation effect of the number of speckle pattern superpositions(equivalent to different exposure times) was compared. The experimental results show that after stacking and averaging 100 speckle patterns with multi-dimensional mechanical disturbance, the standard deviation is only one-thirteenth of that of a single speckle pattern. Finally, the influence of the frequency and amplitude of the scrambler device on the scrambler effect is tested. The results show that higher frequency and amplitude are beneficial to speckle suppression. For the optical fiber in this experiment, the frequency of 1.2 Hz and the amplitude of 6 cm are the most appropriate. Under these conditions, the mean centroid deviation distance of the speckle is the smallest, which is 0.21 pixels.

“Trapiche” tourmaline is a rare variety of tourmaline. Conventional gemstone identification instruments, infrared spectrometers, X-ray fluorescence spectrometers, and hyperspectral imaging techniques were used to test and analyze it. This paper aims to discuss the application of hyperspectral imaging technology in complex samples and enrich the multi-angle study of “trapiche” tourmaline. There are abundant parallel tubes and black solid inclusions in “trapiche” tourmaline. The visual boundary difference is due to the multidirectional distribution of parallel tubular inclusions. The parallel tubular inclusions in region a of the hexagonal column are arranged almost perpendicular to the c-axis. In the r region, the tubular inclusions of the three rhomboid regions divided by the three black arms of “trapiche” radiate outward along the c axis at an angle of 10°～30°. The appearance of the “trapiche” phenomenon comes from the fact that the black inclusions extend from the middle to the outside in three directions, and many of them are attached to the outside or wrapped in the inside of the tubular inclusions. Using the hyperspectral imaging technique, the average reflectance spectral lines and visual images of the region of interest of “trapiche” tourmaline are obtained. The hyperspectral pattern is consistent with the shading of the selection, and a wide absorption band centered at 440 nm in the purple region and 610 nm in the orange-red region appears. The content of Cr and V is high in the combination component analysis. It is considered that the d—d electron transition of Cr3+ causes the green color, and the D-electron spin allowed the transition of V3+. The visualized image shows the test state of the sample along the direction of the c-axis, the edge hexagonal column a gradually disappears, and the central three-square single cone r gradually occupies the entire plane. The pixel ratio between the green zone and the whole plane ranges from 16.81% to 49.96%. The proportion of black arms on both sides of the slice is inconsistent, and the number of black inclusions at both ends of the crystal does not show a linear relationship, but along the +c-axis, the change rate increases. The advantage of hyperspectral imaging technology lies in its ability to rapidly identify and analyze the distribution and concentration of different components in “trapiche” tourmaline and provide detailed information on the mineral’s internal structure and crystal orientation. Identifying microscopic inclusions in minerals is of significant importance for understanding the mineral’s genesis processes and environmental conditions.

In order to study the relationship between the composition and structure of dissolved organic matter (DOM) and methane (CH4) emissions in the paddy soil with Chinese milk vetch returning under different water managements (flooding after a short time delay), we set an incubation experiment, with no green manure and no tillage (DW-0) as control, different water management treatments (immediately flooding, flooding after 5 and 10 days) were set up, three-dimensional fluorescence spectroscopy (3DEEM) and parallel factor analysis (PARAFAC) were used to assess the effects of soil DOM composition and the effect of fluorescence spectra on CH4 emissions. The results showed that both DOM content and cumulative CH4 emissions significantly decreased under G+DW-5 treatment, and both DOM content and cumulative CH4 emissions increased with the extension of dry tillage time. Three fluorescence components, including fulvic-like (C1), humic-like (C2), and protein-like (C3), were obtained by the 3DEEM-PARAFAC method. There were significant differences in the fluorescence components of soil DOM under different water management practices. The highest content of humus-like components (C1+C2) was found in soil DOM under the G+DW-5 treatment, which reached 79.88%, and the restricted principal component analysis indicated that soil DOM components were significantly different under different water management. The DOM Humification Index (HIX) of G+DW-5 was 5.11, significantly higher than other treatments. The correlation analysis showed that CH4 emissions were negatively correlated with the C1 and HIX index but positively correlated with the C2 and BIX index; the results showed that flooding after a short time delay could reduce CH4 emissions by increasing the content of C1 and the level of soil DOM humification. Overall, the composition and structure of DOM in the paddy soil under different water managements are closely related to CH4 emissions, which could provide a scientific basis for reducing CH4 emissions from paddy soil.

The leaf area index (LAI) is an important indicator for characterizing crop growth, so an efficient and accurate estimation of crop LAI can guide field production management. Spectral features can provide information about the reflected and absorbed wavelengths of crops, while texture features can provide information about the gray-scale attributes and spatial location relationships of crops. Previous studies have shown some limitations in estimating crop LAI using only spectral features, and at high LAI levels, the “saturation phenomenon” occurs, resulting in an underestimation of LAI. To fully explore the information of multispectral images from UAVs, texture information of multiple bands was combined to obtain multiband combined texture and to explore whether the fusion of spectral features with multiband combined texture can improve the accuracy of LAI estimation. Firstly, we obtained multispectral data and ground-truthed LAI data of three key fertility stages of potato; then we extracted the texture features of each fertility stage using the gray-level co-occurrence matrix (GLCM) and combined the texture features of multiple bands; then we analyzed the correlation between the vegetation index, the texture features, and the multi-band combination of textures and LAI, and synthesized the correlations and correlations with LAI, and investigated whether the fusion of spectral information and multiband combination of textures could improve the accuracy of LAI estimation. Then, we analyzed the correlation between vegetation index, texture features, and multiband combined texture and LAI and combined the correlation and variance expansion factors to select the preferred vegetation index; finally, we integrated the multiband combined texture and used partial least squares regression (PLSR), ridge regression (RR) and K-nearest neighbors regression (KNR) with parameter tuning to determine the correlation between the vegetation index and LAI, and then used KNR to estimate the correlation between the vegetation index and LAI. KNR will estimate potato LAI at each fertility stage and compare it with the model using only the vegetation index to verify the feasibility of inverting LAI using a multiband combined texture. The results showed that: (1) the correlation between single-band texture, two-band combined texture and three-band combined texture and LAI increased sequentially; (2) the preferred multiband combined texture at each fertility stage of potato showed highly significant correlation with LAI, with correlation coefficients ranging from 0.79 to 0.83; and (3) compared with the model using only the vegetation index, the addition of the multiband combined texture could significantly increase the model’s accuracy and stability. The KNR model had the highest accuracy in estimating potato LAI during the tuber formation period, with a modeling R2 of 0.83，an RMSE of 0.23 m2·m-2, and a validation R2 of 0.75 and an RMSE of 0.25 m2·m-2; the PLSR model had the highest accuracy during the tuber growth period, with a modeling R2 of 0.73 and an RMSE of 0.26 m2·m-2, and a validation R2 of 0.87 and an RMSE of 0.20 m2·m-2; and the PLSR model had the highest accuracy during the starch accumulation period, with a modeling R2 of 0.73 and an RMSE of 0.26 m2·m-2; and the PLSR model had the highest accuracy during the starch accumulation period. The PLSR model had the highest estimation accuracy, with modeling R2 of 0.73 and RMSE of 0.31 m2·m-2, and validation R2 of 0.84 and RMSE of 0.25 m2·m-2. This method can provide a reference for the UAV multispectral combination of texture features to estimate potato LAI.

To achieve timely monitoring of plant growth, rapid detection of differences in the distribution of catalase activity in melon leaves under different light intensities is essential. In this study, melon leaves were treated with different light intensities. Then the leaves were scanned using fluorescence hyperspectral imaging to extract the average spectral reflectance of 300 leaf samples, and the raw spectra were pre-processed and optimised by four pre-processing methods. Using interval Variable Iterative Space Shrinkage Approach (iVISSA), Competitive adaptive reweighted sampling (CARS), Genetic algorithm partial least squares algorithm(GAPLS), Iterative retained Information Variable(IRIV), and Variables Combination Population Analysis(VCPA) were used to extract the feature wavelengths. The partial-least-squares regression (PLSR) model screened the optimal feature wavelengths. Based on the preferred feature wavelengths, Principal component regression (PCR) model, Multiple linear regression (MLR) model, Convolutional Neural Network (CNN) model, Least Squares Support Vector Machine (LSSVM) model, and the results show that Baseline-IRIV-MLR model has the highest recognition accuracy, with an accuracy of 0.852 in both training and prediction sets. The results of this study provide a theoretical basis for applying fluorescence hyperspectral imaging technology in the quality evaluation of melon crops and technical support for the development of precision agriculture.

With the rapid growth of the economic level, people’s demand for mahogany products is increasing day by day. As precious mahogany, criminals replace sandalwood and counterfeit it because of its spiritual pursuit and high price. To maintain the order of the timber market and the interests of consumers and realize the rapid non-destructive detection and identification of sandalwood, it is necessary to establish a fast and reliable intelligent identification method of sandalwood. This paper used near-infrared spectroscopy to extract the spectral information of Pterocarpus santalinus and its similar blood sandalwood. The qualitative analysis method, partial least squares discriminant analysis ( PLS-DA ), and error back propagation artificial neural network ( BPNN ) were used to establish the calibration model of spectral information. Then, Pterocarpus santalinus and its similar wood blood sandalwood were identified. By analyzing and comparing the advantages and disadvantages and recognition accuracy of these three models for these two kinds of wood recognition, the feasibility of this method in the recognition of sandalwood and sandalwood is verified. The experimental results show that the three discriminant models can quickly identify the wood spectral images, and can quickly and non-destructively classify and identify sandalwood and sandalwood. In the case of selecting different image processing methods, the results of the three discriminant models are different, and the recognition results are also different. Further analysis reveals that when the spectral data are modeled using the BPNN model, the peak and trough eigenvalues of the original spectral data within the wavelength range of 866~2 533 nm are utilized as input following preprocessing. It is observed that when the number of input layer nodes is Setto 24 eigenvalues and the hidden layer consists of 13 neurons, the model achieves the smallest root mean square error, with an accuracy rate of 96.43%. Additionally, shortening the spectral range does not result in an improvement in the model’s recognition rate. The BPNN model demonstrates the highest recognition rate across the full band range among the three models. The experimental results indicate that the combination of artificial neural network modeling and near-infrared spectral feature extraction technology yields a high recognition accuracy in identifying sandalwood wood, surpassing trevious methodologies. This study contributes to mitigating the subjectivity inherent in manual recognition processes. The utilization of computers can expedite the recognition process, while the enhanced accuracy aids in maintaining order within the wood market and safeguarding consumer rights. Simultaneously, it offers insights into realizing intelligent visual recognition of sandalwood and furnishes technical support for the sustainable development of the mahogany industry.

Oil-paper insulation equipment is one of the key components in power systems, and accurately assessing the aging status of electrical equipment’s oil-paper insulation is crucial for ensuring the safety of power grids. Acetone is an important indicator of the aging process of insulation paper, and the rapid, sensitive, and accurate determination of acetone content dissolved in insulation oil is of significant importance in evaluating the aging of oil-paper insulation systems. In this study, a surface-enhanced Raman scattering (SERS) substrate based on silver nanowires (AgNWs) decorated with gold nanostars (AuNSts) was successfully prepared on a silicon-gold (Si-Au) film surface using the ion sputtering method. This substrate was utilized for rapid assessment of acetone content in insulation oil. Morphological characterization of the substrate by scanning electron microscopy revealed that the dual-metal satellite-surrounded structure formed by smaller gold nanoparticles sputtered onto the larger surface area of AgNWs not only effectively enhanced surface plasmon resonance but also provided a protective barrier layer for AgNWs, significantly improving the overall antioxidant performance of the substrate. SERS detection results showed that compared to Si-Au, Si-Au-AgNWs, and Si-AgNWs-AgNts substrates, the Si-Au-AgNWs-AuNSts substrate exhibited superior SERS performance, with the lowest detection limit of 40 mg·L-1 for acetone dissolved in insulation oil after water (H2O) extraction. Moreover, the substrate demonstrated high consistency and stability, with an RSD value of 4.87%, and after 30 days, the target peak signal only decreased by 6.14%. This method provides an effective approach for achieving reliable on-site detection of acetone dissolved in oil.

The development of observation technology has led to massive spectral data. How to automatically classify these data has received attention from researchers, the most important of which is feature extraction. Given the limitations of manual processing, most of the research uses machine learning algorithms to extract feature-based spectral data. However, these machine learning algorithms cannot handle massive spectral data due to the high spatial and temporal complexities. The pre-trained models emerging in recent years have excellent feature extraction capabilities. Still, there is little research on the effectiveness of such a model in the feature extraction of spectral data. Therefore, this paper takes the stellar spectral data as the research object separately introduces the pre-training models such as BERT, ALBERT, GTP, and Convolutional Neural Networks (CNN) for feature extraction and classification of the stellar spectral data, and tries to verify the effectiveness of these pre-training models for feature extraction of stellar spectral data by comparing the experimental results. Python programming language is used to write the spectral classification program. Based on the feature extraction of the pre-trained models, the CNN model in TensorFlow 1.14 is utilized for spectral data classification. The dataset used for the experiment is the SDSS DR10 stellar spectral dataset, including K-type, F-type, and G-type. The grid search and 5-fold cross-validation are utilized to obtain the experimental optimal parameters. The BERT model has the highest classification accuracies compared to ALBERT and GPT with the same experimental conditions. In terms of the average classification accuracies, the average classification accuracies of the BERT model are 0.025 1, 0.021 5, and 0.022 5 higher than that of ALBERT, and 0.049 7, 0.042 4, and 0.043 2 higher than that of GPT, on the K-type, F-type, and G-type stellar datasets. It is easy to draw the following conclusions by analyzing the experimental results: Firstly, the classification accuracies improve with the scale increase of training data; Secondly, the same model has the highest classification accuracies on the same training dataset of K-type stellar, followed by the F-type and the G-type; Thirdly, the BERT model has the best ability of feature extraction compared with ALBERT and GPT.

The content of ancient calligraphy artifacts contains crucial information documenting the social civilization of ancient China. Obtaining and identifying hidden inscriptions within the paper surface is important for historical background research on cultural relics and the collection of humanistic classics. However, traditional methods for identifying hidden inscriptions in calligraphy artifacts rely heavily on manual interpretation, which requires extensive expertise from researchers and results in a time-consuming analysis that may inadvertently cause secondary damage to the artifacts. To address this challenge, hyperspectral remote sensing, with its non-contact and efficient characteristics, captures the spatial properties of paper-based artifacts and acquires rich spectral information. This enables the digitization and storage of calligraphy artifacts. Initially, the Minimum Noise Fraction (MNF) transformation technique was utilized to reveal latent blurry information in calligraphy artifacts. Subsequently, a spectral transformation method based on Linear Difference Enhancement (LDE) was developed to identify these details further, and statistical analysis was conducted on the spectral parameters before and after transformation and the component images extracted by MNF. By utilizing entropy evaluation, we obtained the most information-rich spectral image, ultimately enabling the successful identification of the hidden inscriptions within the calligraphy artifact. The research results demonstrate the following: (1) The MNF transformation of the hyperspectral image of the calligraphy artifact reveals the blurred patterns hidden in the inscriptions. These patterns exhibit similarities in spectral morphology with other content elements of the calligraphy artifact but with differences in reflectivity. (2) The LDE algorithm effectively amplifies the relative differences between hidden information within the calligraphy artifact and the spectral bands of the inscriptions. LDE significantly enhances most spectral features of the hidden inscriptions after enhancement. (3) Following LDE processing, the calligraphy artifact data shows improved image entropy values in the wavelength range, spectral characteristics, and MNF sub-components. Particularly, the entropy value of the spectral variance (SV) image after LDE processing reaches 6.74 bits. (4) After LDE processing, the SV image of the calligraphy artifact successfully identifies the hidden inscriptions on the paper that do not belong to the “Heart Sutra” content. This finding proves that the paper used for this calligraphy artifact is dedicated to sutra writing. This discovery effectively reveals and identifies hidden inscriptions behind Emperor Qianlong’s ink treasure, enriching the historical and humanistic background of the artifact. It also provides scientific, theoretical, and technical support for future studies on extracting hidden inscription information in ancient calligraphy artifacts.

Iron ore grade is an important index used to evaluate the degree of wealth and economic value of iron ore, and the verification efficiency of iron ore grade greatly influences the efficiency of iron ore mining. Because of the advantages of hyperspectral images in the fields of substance classification and content inversion, such as fast analysis speed, high accuracy, and non-destructive, this study collected hyperspectral images of Anshan-type iron ore in the two bands of VIS-SWIR and NIR, respectively, and discussed the feasibility of realizing grade inversion of Anshan type iron ore based on hyperspectral images. First, the average spectral representation in the ROI of the hyperspectral image is extracted, and the spectral data of the corresponding samples are transformed by multivariate scattering correction (MSC) and Standard normal variate transformation (SNV), respectively. Then, Monte Carlo uninformative variable elimination (MCUVE), competitive adaptive reweighted sampling (CARS), and successive projections algorithm (SPA) were used to extract the characteristic bands of the spectral data before and after the transformation. Finally, the quantitative inversion model of Anshan type iron ore’s iron grade is established using radial basis function neural network (RBFNN) and extreme learning machine (ELM). The results show that after the MSC transformation of spectral data in the VIS-SWIR range, the ELM grade inversion model established by using the feature bands extracted by CARS has the best effect (R2=0.90, RPD=3.02, RMSE=3.27, MAE=2.77). Applying the MSC-CARS-ELM model to the VIS-SWIR hyperspectral image of an Anshan-type iron ore sample can generate a pixel-level iron ore grade distribution map. This study provides a new method for realizing grade inversion and visualization of Anshan-type iron ore quickly and effectively, which has important application value in geology and mining.

The mid-infrared(MIR) atmospheric strong absorption wavelength bands(near 2.7 and 4.3 m) have the characteristics of suppressing background clutter signals, which are often used on the payload of infrared early warning satellites to achieve stable detection and tracking of targets. The background clutter fluctuation level has an important impact on MIR target detection, and it is of great significance to study the spatial domain fluctuation of surface-atmosphere clutter near the 2.7 and 4.3 m bands. First, the MODTRAN radiative transfer model is used to iteratively simulate the optical depth (OD) of water vapor and CO2 under six standard atmospheric models, screen the strong absorption bands according to the criterion of OD>1, and intersect the band calculation results under various atmospheric models. The final strong absorption bands were 2.52～2.83 and 4.18～4.47 m. Then, global monthly averaged surface and atmosphere background products are produced based on remote sensing satellite data products and data assimilation information. Among them, combined with the monthly average surface reflectance/emissivity products (MYD09A1/ MOD11C3) of MODIS shortwave infrared and mid-infrared bands (2.13, 3.75, 3.96 and 4.05 m), the non-negative matrix factorization (NMF) method was used to reconstruct global surface reflectance/emissivity products at strongly absorbing bands; Produce monthly-averaged global surface temperature products based on ERA5 reanalysis data; The global monthly average products of water vapor and cloud optical thickness (COT) are obtained by splicing and fusing MODIS atmospheric products (MOD05~~L2 and MOD06~~L2) respectively; The CO2 global monthly average product comes from the data assimilation product of the OCO-2 satellite; In order to spatially match various data products and save computing resources, global products are limited to the 60° north and south latitude range and resampled to a spatial resolution of 1°×1°; Next, under cloudy and cloud-free conditions, the background clutter intensity of two strong absorption bands was simulated and calculated pixel by pixel, and the spatial distribution pattern of clutter was analyzed. Finally, the 11×11 square window neighborhood statistics method was used to calculate the clutter spatial domain fluctuation in the simulation results, convert the clutter fluctuation under 95% probability, and conduct histogram statistics. From the perspective of clutter suppression, a comparison of infrared target detection performance in two absorption bands is given. Study results show that under cloud conditions, the fluctuation level of background clutter in the 2.52～2.83 m band has the characteristics of “a point-like peak, regional enhancement, and overall low values”. In contrast, the clutter fluctuation level in the 4.18～4.47 m band shows the characteristics of “regional high values, patchy enhancement, and low values at high latitudes”. Under cloud-free conditions, the background clutter fluctuation level in the 2.52～2.83 m band increases significantly in the low water vapor content area. In contrast, the clutter fluctuation level in the 4.18～4.47 m band is generally low, and most of them are controlled within 2×10-4 W/m2/sr/m. On the global scale, the infrared target detection performance is better in the 2.52～2.83 m band when there are clouds and in the 4.18～4.47 m band when there are no clouds. The results of this study can provide a reference basis for detecting infrared targets in terms of spatial domain law and spectral band optimization, which is of great value for enhancing the detectability of targets.

Chlorophyll is a crucial indicator for assessing grasslands’ photosynthetic capacity and physiological condition.With its rich spectral information, hyperspectral remote sensing has become an important means for non-invasively estimating chlorophyll content in grasslands. However, there is a scale mismatch between the canopy hyperspectral data and the measured leaf chlorophyll values, leading to hyperspectral chlorophyll retrieval’s low accuracy. Therefore, this paper proposes a hyperspectral retrieval method for natural grassland Canopy Chlorophyll based on the green cover rate. The typical natural grassland in Hulunbuir, Inner Mongolia, was selected as the research object. The measured leaf chlorophyll relative content values were obtained by ASD hyperspectral spectrometer, SPAD chlorophyll meter, and mobile phone digital photos.The results indicate that the correlation between vegetation indices and SPAD ranges from -0.74 to 0.76, which is generally higher than the average correlation of SPAD pushed up from -0.63 to 0.50. Green cover media pushed the measured values of leaf chlorophyll relative content to the sample canopy scale. First derivative spectra and 42 common chlorophyll spectral indices were used to construct a hyperspectral retrieval model (SPAD) of natural grassland Canopy Chlorophyll based on green cover rate~~ cover. The single variable optimal grassland Canopy Chlorophyll retrieval modelR2=0.689, RMSE=2.714, RPD=1.752; The best regression model of grassland Canopy Chlorophyll wasR2=0.833, RMSE=2.019, RPD=2.354. The results show that the hyperspectral retrieval accuracy of chlorophyll content in natural grassland canopy can be effectively improved by extrapolating the measured value of chlorophyll content in grassland leaves to the canopy scale based on the green cover rate.

Paper-based cultural relics hold crucial significance to the historical and cultural heritage and the continuation of the national spirit in China. At the same time, the aging of paper seriously affects the longevity of artifacts. Paper viscosity is an important indicator reflecting the degree of paper aging. Traditional paper viscosity measurement methods are destructive experiments that cause inevitable secondary damage to valuable cultural relics. Hyperspectral remote sensing technology can achieve non-destructive and rapid analysis to address this issue, providing an effective approach to establishing a paper viscosity inversion model. In this study, drying and moist-heat paper aging were taken as the research subjects, and110 groups of simulated paper aging samples with measured viscosity content were obtained. Through spectral filtering, a hyperspectral database of paper aging was established. Based on this, nine data processing methods, including original spectrum first-order derivative, original spectrum second-order derivative, original spectrum reciprocal logarithm, original spectrum reciprocal logarithm first-order derivative, multiplicative scatter correction, multiplicative scatter correction spectrum first-order derivative, multiplicative scatter correction spectrum second-order derivative, multiplicative scatter correction spectrum reciprocal logarithm, and continuous wavelet transform were analyzed, as well as two feature selection methods, competitive adaptative reweighted sampling (CARS) and correlation coefficients (R), were analyzed in combination as input for the models to identify the best model through dataset partitioning validation. Research results have shown: (1) In the original spectrum, the correlation at 430 nm is the highest with viscosity, with an R-value of 0.75. After data processing, the R-value at 430 nm increases to 0.874 following the application ofthe reciprocal logarithm first-order derivative of the original spectrum method, which significantly enhances the paper viscosity information in the spectrum; (2) After the original spectrum is transformed into the second-order derivative, the correlation at 578 nm is the highest, with an R-value of 0.57, significantly lower than the highest R-value (0.75) in the original spectral data. This finding indicates that the second-order derivative is unsuitable for paper viscosity estimation. Meanwhile, the maximum correlation coefficients of other transformation methods are all higher than the original spectrum, demonstrating the effectiveness of spectral transformation; (3) For the original spectrum, as the decomposition scale increases, the paper viscosity information contained in the spectrum decreases, and the highest correlation coefficient occurs at 484nm at the decomposition scale of 2, with a value of 0.873; (4) Among different input combinations, the model with the highest accuracy uses the reciprocal logarithm first-order derivative of the original spectrum as input. Based on the CARS method, the support vector regression (SVR),random forest (RF), and AdaBoost models have R2 values of 0.96, 0.93, and 0.93, respectively, and RMSE values of 14.80, 18.33, and 19.79 mL·g-1 for the validation dataset. We recommend prioritizing using the reciprocal logarithm first-order derivative of the original spectrum as inputs and the SVR model with the CARS feature selection algorithm for paper viscosity inversion. The above research findings demonstrate the applicability of hyperspectral technology for non-destructive analysis of paper viscosity, providing a scientific basis for the restoration work of paper-based cultural relics.

Referring to the current use of sparse representation algorithms to extract camouflaged targets from hyperspectral images, the selection of the background dictionary is affected by the “same spectrum of different objects” of the hidden targets, resulting in the inability to detect camouflaged targets accurately. In this paper, we take the grassland camouflage net and desert camouflage net as the research objects, collect the visible-near-infrared reflectance spectra of the camouflage net and airborne hyperspectral images respectively, and analyze the spectral characteristics of the background pixels and camouflage target pixels in the camouflage net and airborne images measured outdoors. Taking advantage of the fact that the spectra of the camouflage nets measured outdoors and the background in the airborne images are different and the possibility of neighboring image elements belonging to the same feature is high, the background dictionary selection method based on the constraints of Euclidean distances and image homogeneity features is proposed, and the sparse representation of the background dictionary is further utilized to identify the target. The results show that (1) in the wavelength range of 750~1 000 nm for grass camouflage, the reflectance of the background pixel spectrum in the image is higher than that of the camouflage net spectrum. For desert camouflage: in the range of 550～700 nm, the reflectance of the background pixel spectrum in the image is higher than that of the camouflage net spectrum. (2) By establishing spatial and spectral feature constraints with the maximum spectral Euclidean distance to the target image element and the highest homogeneity with neighboring image elements, 413 background image elements in the grass camouflage image and 507 background image elements in the desert camouflage image were selected as the background dictionary. (3) Based on the improved background dictionary selection method, the sparse representation algorithm is utilized to identify the camouflage targets, and the results can accurately discriminate the location and number of camouflage targets. The area under the curve (AUC) of the receiver operating characteristics for the detection of grass camouflage targets and desert camouflage targets reaches 0.96 and 0.98, respectively, indicating that the algorithm has good detection performance for both grass camouflage targets and desert camouflage targets.

To achieve wide-scale, accurate, and timely monitoring of pear nutrients during the growth and fruiting periods of pear trees and to further provide protection and adjustment strategies for pear fertilization management and pear fruit quality improvement, we used the ASD FieldSpec3 hyperspectral spectroscopy system to monitor the nutrient content of pear trees. The ASD FieldSpec3 hyperspectrometer was used to obtain leaf hyperspectral data during the fruit expansion and ripening periods of pear trees and to collect information on the leaf nitrogen, phosphorus, and potassium content. The effects of spectral preprocessing methods such as raw reflectance, first-order derivative transformation (FD), convolutional smoothing algorithm (SG), and standard normal variable transformation (SNV) on the fitting effect of the hyperspectral reflectance monitoring model were comparatively analyzed through the raw spectral curves. Principal component analysis (PCA), competitive adaptive reweighted sampling (CARS), and successive projection algorithm (SPA) are then used to select the characteristic bands of hyperspectral data. After that, partial least squares regression (PLSR), support vector regression (SVR), random forest (RF), gradient boosted tree (GBDT), convolutional neural network (CNN), and deep forest (DF) algorithms were used to establish monitoring models based on the edge bands to screen the optimal hyperspectral monitoring models for the three nutrient elements, nitrogen, phosphorus, and potassium, in pear trees. The optimal modeling combination of nitrogen was the DF regression model of PCA eigenbands pre-processed by SG-SNV (R2=0.928 3, RMSE=0.238 1 g·kg-1). The optimal modeling combination of phosphorus was the GBDT regression model of SPA eigenbands pre-processed by SG-SNV (R2=0.936 7, RMSE=0.043 1 g·kg-1). The best modeling combination for potassium was the DF regression model (R2=0.954 4, RMSE=0.276 7 g·kg-1) for the SG-SNV preprocessed PCA feature band. The monitoring model based on hyperspectral edge bands was well fitted (R2>0.9), which can realize the accurate monitoring of nitrogen, phosphorus, and potassium content in pear fruit during expansion and ripening.

The plastic layer is an important intermediate phase in the coking process. Its structure evolution and the interference between different structures in the coal blending process are of great significance for understanding coal’s properties and the blending’s interaction mechanism. This paper used self-made coking-related properties measuring device to conduct coking experiments on different blend ratios. The evolution products of the plastic layer were obtained by the interrupted cooling method, including the softening zone (SFZ), melting zone (MPBZ), flow zone (MZ), and resolidification zone (RSZ). FTIR determined the raw materials that were obtained. FTIR was divided into four bands of 3 600～3 000, 3 000～2 800, 1 800～1 000 and 900～700 cm-1 for peak fitting analysis. I-aromatization, DOC-polycondensation degree, —CH2— aliphatic structure, A’-hydrocarbon generation capacity, and C-oxygen-containing functional group were used to explore the changes in the coking process and the interaction among the same evolution products of blends with different mass ratios. The results show that I and DOC increased gradually in the special temperature field, and the changes were particularly significant from MZ to RSZ. With the release of volatiles and tar, the aliphatic group —CH2 and A’ as a whole showed a downward trend. The content of C fluctuates slightly in MPBZ and MZ, but the overall trend is also downward. The condensation of aromatic carbon structure induces the decomposition of aliphatic structure and oxygen-containing functional groups. MZ to RSZ in the plastic layer is an important part of coking process. The I1 and DOC of the same evolution products of mixed coals with different mass ratios have good additive properties in SFZ, MPBZ, and MZ, with the degree of fitting R2 reaching 0.744, 0.71, 0.775 and 0.74, 0.266, 0.773 respectively. The interaction of other structures is influenced by many factors of the process of pyrolysis and bonding, so these do not have additivity. Therefore, the aromatic carbon structure at the resolidification zone and the lack of additivity of the aliphatic structure and heteroatoms are important factors affecting the different properties of coke.

Coal is very sensitive to geological environmental conditions such as temperature and pressure. Driven by tectonic stress, coal’s “graphite crystallite” structure grows in orientation, and the physical chemistry and structure show anisotropy. X-ray diffraction and Raman spectroscopy characterized the samples with different degrees of graphitization. The results show that the coal seam slides along the layer under high temperatures and shear stress. The structural deformation makes the graphite crystallites rotate and preferentially oriented, increasing the stacking height along the c-axis direction. The anthracite stage is divided by the stacking degree Lc≤5 nm, which is a turbulent layer structure with random orientation or irregular arrangement of carbon layers, optically isotropic; when Lc≥30 nm, it is regarded as a sign of the formation of a perfect graphite structure, and the optical anisotropy is significant. It belongs to the transition state structure (semi-graphite stage) with an imperfect graphitization structure between 10~20 nm. The 1 350 cm-1 band (D1) and 1 620 cm-1 band (D2) in the Raman spectra of SXL100 and SXL130 samples in the graphite stage are obvious. Still, the full width at half maximum of the D1 to G peak (ID1/IG, R1), and the intensity ratio of D2 to G peak (ID2/IG, R3) in the Raman spectra of the graphite edge plane are significantly higher than those of the preferred orientation plane, indicating that the Raman parameters such as intensity ratio depend on the orientation of the edge plane of coal-based graphite. The intensity of the D1 peak depends on the degree of defect or disorder of the sample. The D2 peak of the edge plane has asymmetric characteristics, and the bimodal structure is significant. The D′1 peak changes with the D2 peak, which also shows the spectral behavior of the edge plane defect. The stages of anthracite (R1≥1.0), semi-graphite (1.0>R1≥0.5), and graphite (R1<0.5) were divided by the defect density or order degree index R1 of coal measures graphite, and the uniformity of coal measures graphite and the proportion and distribution of components with different graphitization degrees were evaluated. It was found that the proportion of advanced evolution to semi-graphite structure in the metamorphic anthracite CM130N sample was 3.52%, the proportion of anthracite structure in the semi-graphite BC210 sample was 46.40%, and the SXL130 sample was graphitized as a whole. However, the proportion of anthracite structure was still 3.84%, and there were still defects in the preferred orientation plane and edge plane structure. The established method has more advantages in distinguishing anthracite and semi-graphite. When the incident direction of the Raman laser is constant, R1 and R3 parameters can be used to explore the Raman spectral characteristics of coal measures graphite preferred orientation plane affected by tectonic stress and to evaluate the heterogeneity of graphitization and the orientation of graphite crystallites.

With the rapid socioeconomic development in theLake Taihu basin, a large amount of industrial, agricultural, and domestic wastewater has been discharged into Lake Taihu, and nutrients have been enriched, especially in the northern lake regions of Lake Taihu, where algal blooms occur frequently. In the spring, summer, and fall of 2015, we conducted six field sampling campaigns at 24 sites in northern Lake Taihu’s high-incidence area of algal blooms. Absorbance coupled with three-dimensional fluorescence and parallel factor analysis (3DEEMs-PARAFAC), together with water quality parameters, were used to investigate the spatial and temporal variability of DOM and water quality in Lake Taihu under the influence of algal bloom to provide scientific support for the management and restoration of the Lake Taihu ecosystem and ensure the safety of its water supply. During the sampling period (from May 12 to October 20), the three lake regions experienced significant algal bloom and degradation. The algal blooms were closely related to the input of exogenous nutrients. During the summer, the rapid growth of algal biomass significantly consumed NO3--N generated NO4+-N and inhibited the denitrification process of microorganisms. Parallel factor analysis identified four fluorescent DOM components: two protein-like components (tryptophan-like C1 and tyrosine-like C3) and two humic substances (microbial humic-like C2 and terrestrial humic-like C4).Terrestrial inputs from the northwest significantly affected the composition and structure of DOM in northern Lake Taihu, showing a trend of decreasing abundance and terrestrial aromaticity from west to east. In addition, the large amount of terrestrial DOM also provided an important carbon source for the growth of planktonic algae, which further influenced the composition and structure of DOM and led to the increased participation of autochthonous DOM (C2 and C3) with strong microbial activity in the lake carbon cycling process. There were differences among the three northern algal areas, and the algal bloom management in these three lakes needs to be adapted to local conditions; exogenous inputs should be reduced in the northwestern part of the lake, algal bloom monitoring and ecological restoration should be strengthened in Meiliang Bay, and ecological diversity should be preserved and restored in the northeastern part of the lake to enhance the natural purification capacity.

Quinolone antibiotics (QNs) are extensively utilized in treating diseases and animal husbandry owing to their potent antimicrobial properties. However, the excessive utilization of QNs and their release into environmental water sources via wastewater results in their accumulation in natural aquatic ecosystems. The consequence of this phenomenon is the extensive propagation of bacteria that are resistant to antibiotics, as well as the proliferation of resistance genes in aquatic ecosystems. This poses a significant risk to both environmental ecology and human well-being. Traditional techniques employed in detecting QNs exhibit notable sensitivity and accuracy. However, these methods are characterized by their time-consuming nature, reliance on costly equipment, and the inherent difficulty of conducting on-site assessments. Fluorescence analysis technology necessitates a brief detection time, particularly in the case of three-dimensional fluorescence spectroscopy, which enables the acquisition of a substantial amount of target information within a limited timeframe. By integrating data statistics and machine learning models, mathematical methods can efficiently identify multiple QNs. In this study, the fluorescence spectroscopic information of QNs was extensively employed, and the support vector machine regression (SVMR) algorithm was utilized to develop prediction models for QNs, specifically ofloxacin (OFL) and norfloxacin (NOR). The fluorescence spectroscopic data of the unknown samples was subsequently fed into the developed models to obtain the determination outcomes efficiently. While constructing the model, a comparison was made between two supervised learning methods, i.e., PLS-DA and SVMR. It has been determined that SVMR exhibits a strong predictive capability. By manipulating parameters and kernel functions, it was possible to achieve a good linear range for determining OFL and NOR, spanning from 2 to 600 g·L-1. The resulting linear correlation coefficients exceeded 0.992 0, and the detection limits were 0.064～0.080 g·L-1. The validated method was proven applicable to real water samples, i.e., the recoveries were 98.62%～104.01% for seawater and 103.90%～105.89% for reservoir water. This method offers the advantage of rapid detection speed, allowing for the completion of quantitative analysis of an unknown sample within a mere 3 min. This rapid screening process identifies potential risk factors associated with QNs in the environment. This study employs a novel approach by integrating SVMR with fluorescence spectroscopy to develop a rapid detection method for QNs in real water samples. The proposed method offers a new and scientifically reliable solution for rapidly detecting QNs in environmental water.