This paper carried out CO2 inversion experiments based on the remote sensing data from the greenhouse gases monitoring instrument (GMI) on the GF-5 satellite in China, calculated the CO2 profile samples according to the differences in China’s regional characteristics, and constructed the representative sample set suitable for China’s regional characteristics. Then, it substituted the CO2 profile obtained by statistical inversion as the initial value into the physical inversion method to form a new algorithm for synergistic statistics and physical methods. By analyzing the inversion results of the new algorithm, we conclude that the collaborative inversion algorithm improves the accuracy by 47.7% on the basis of using the physical inversion algorithm alone, and the correlation between the inversion results of the new algorithm and the observation results provided by the international satellite of the same type, OCO-2, reaches 88.5%.

In this paper, an underwater image restoration method based on non-uniform incident light imaging model is proposed to solve the problems of blurring, low contrast, and color distortion of underwater images. Through wavelet decomposition, we can suppress the scattering light in the lowest frequency subband while reducing the noise and enhancing the details in high-frequency subbands. The contrast deviation and the color deviation caused by underwater non-uniform incident light can be corrected by dynamic range stretching and histogram matching. The experimental results show that the proposed method can restore images with high definition, balanced contrast, and natural colors. The average of the underwater color image quality evaluation (UCIQE) index is 0.6240, which is superior to that of the algorithms reported in the references.

Due to their low costs, low power consumption, and high speed, passive optical interconnection data centers have been widely concerned. In a passive optical interconnection data center, the channel number available at the same time cannot exceed the available wavelength number, and the wavelength number limits the total throughput. To reduce the wavelength requirements for optical interconnection data centers and improve the wavelength utilization and scalability of the optical interconnection data centers, we apply NRZ+Manchester signals combined with polarization multiplexing to the optical interconnection data centers in this paper. Four servers share one wavelength, which increases the throughput by four times. Simulation results show that the proposed scheme can realize the optical interconnection of 64 servers in the rack at the rate of 10 Gbit/s, where only 16 wavelengths are needed. At the rate of 25 Gbit/s, only eight wavelengths are required for the optical interconnection of 32 servers. This scheme can save 75% wavelengths, reduce equipment costs, and improve the throughput of optical interconnection data centers.

To achieve high-accuracy temperature sensing, this paper designed a fiber temperature sensor with a composite structure based on the vernier effect and aluminum alloy substrate sensitization. The sensor is composed of a cascaded Fabry-Perot interferometer (FPI) formed by splicing a hollow-core fiber and two single-mode fibers, a fiber Bragg grating (FBG) in series connection with the cascaded FPI, and an aluminum alloy substrate. Its characteristics such as reflection spectra, transmission field, and thermodynamics are expounded by the three-beam interference theory, the beam propagation method, and finite element analysis. By controlling the geometric lengths of the hollow-core fiber and the single-mode fiber, we can flexibly adjust the vernier effect sensitization multiplier and temperature measurement resolution. The composite structure was designed to measure not only small temperature changes but also the absolute values of temperatures. The experimental results show that the temperature sensitivity of the cascaded FPI and FBG based on the vernier effect is 138.4 pm/℃ and 37.4 pm/℃, respectively, and the temperature repeatability and fast response of the sensor are good. The designed sensor can be widely applied to high-precision measurements.

An all-optical fiber sensing system based on fluorescent fiber is proposed to detect partial discharge faults during ground simulation experiment for the high-voltage supply system in the cabin of a near-space vehicle. The partial discharge response parameters in the equivalent vacuum state of near space are obtained, and the equivalent ground height of the proposed system is more than 20 km. A partial discharge fault in the vacuum cabin of the flight vehicle with a diameter of 86 cm can be detected. Experimental results confirm that the proposed all-optical sensing system has suitable optical electron response and sensing characteristics for partial discharge faults at different pressures.

Solar energy calculation aims at calculating solar radiation collected by a solar receiver using solar geometry and radiation data. The essential calculation includes the incident angle of solar rays on the receivers and the projection angle of solar rays on a specific plane, although the calculation procedure quite differs due to the receiver differences in structure, orientation as well as optical characteristics. To date, the spherical coordinate system is taken as the basis for describing the law of solar motion and related calculation. However, such a conventional method makes the calculation process complex. In this article, taking the fixed and tracked solar panels as case studies, it is analyzed in detail that how to use vector algebra for coordinate transformation, how to select and establish coordinate systems, and how to use vector algebra to analyze the spatial angle relationship between line and line (plane) as well as the spatial transmission law of solar rays after mirror reflection. The analysis results show that the calculation process of the incident angle of solar rays on fixed and tracked solar panels as well as the projection angle of solar rays on a specific plane can be greatly simplified by selecting and establishing a reasonable coordinate system according to actual needs, and vector algebra is very suitable for analyzing the three-dimensional transmission process of solar rays within a reflective linear solar concentrator.

This paper proposes an optimal design method of filter window for the holographic three-dimensional (3D) display optical system. First, based on scalar diffraction theory, the complex amplitude coding method is used to obtain the kinoform, and the spectrum distribution of the reconstructed light field of the kinoform on the focal plane of the optical system is studied. Then, the spectrum center position and frequency distribution formula of the reconstructed object light field on the focal plane of the optical system are derived and verified by experiments. Finally, an optimized design method for the filter window of the holographic display system based on the phase-only spatial light modulator is proposed. The results of theoretical simulation and optical experiment show that the optimization method can obtain kinoform with optimal encoding, and can design a high quality holographic 3D display system to obtain high quality holographic reproduced images.

To improve the accuracy of gun image stabilization calibration, a fitting calibration model was designed to improve the calibration property of a gun's collimator, which is expected to increase the shell's hitting probability. A threshold segmentation model fused via frequency-domain filtering was constructed to eliminate the influence of uneven imaging gray levels. A long line-fitting model and calibration-positioning axis model were constructed to calibrate the target center and realize accurate detection of the image's stabilization calibration, respectively. The results demonstrate that the proposed method exhibits good segmentation and denoising effect and can effectively eliminate the uneven phenomenon of imaging gray levels. The fitting correlation coefficient of the horizontal and vertical demarcated long lines is 0.995646 and 0.993913, respectively, and the calibration error of horizontal and vertical coordinates is less than or equal to 0.15% and 0.005%, respectively. Overall, the experimental results demonstrate that the designed calibration method can accurately achieve calibration detection of gun sight.

Owing to the spatiotemporal randomness of mitosis, the automatic identification and accurate location of mitosis in living cells are challenging tasks for researchers. Herein, a deep learning-based detection method was proposed to automatically identify and locate mitosis in living cells. Here, we built a deep neural network called DetectNet by improving the backbone network of YOLOv3 and introducing an attention mechanism. Under the condition of bright-field microscopic imaging, multiscale images of living cells were acquired and then a dataset was constructed to train the network. The trained network DetectNet was compared with multiple object detection algorithms, and its effectiveness was verified. Experimental results show that aiming at the bright-field microscopic images, DetectNet can directly identify and locate mitosis from the multiscale live cell images with a large field, achieving a higher detection accuracy and faster detection speed compared with other multiple object detection algorithms. Thus, DetectNet shows a great potential application value in the fields of biology and medicine.

With the high-precision nano-vibrator as the vibration target to be measured, this paper studied the swept-source optical coherence tomography (SS-OCT) method based on ultra-small gradient index (GRIN) fiber probes for vibration measurement. The working performance of the integrated SS-OCT based on ultra-small GRIN fiber probe was assessed. In theory, the system can measure the maximum peak-to-peak value of vibration up to 12.5 mm and the maximum frequency up to 25 kHz. The corresponding vibration measurement system was developed and the tests and analysis of micro-vibrations were performed. The results show that, with the peak-to-peak value range of 1 nm--5 μm for the nano-vibrator and the frequency range of 1--200 Hz, the integrated SS-OCT vibration measurement system can detect micro-vibrations with a frequency of 1 Hz and a peak-to-peak value of 6 nm or above. For a single-frequency sinusoidal vibration signal with an amplitude of 25 nm and a frequency of 10 Hz, the system has a repeatability-test standard deviation of 0.003. It shows that the integrated SS-OCT system based on the ultra-small GRIN fiber probe has the feasibility of measuring micro-vibrations, which provides an experimental basis for further research on its application in precise measurement fields such as micro-vibrations and micro-displacements.

Single pixel detection technology can realize two-dimensional imaging of object reflectivity or transmittance, but the imaging results will be affected by the incident angle of the illumination light, the shape of the object surface and the spatial orientation of the detector surface. The research shows that the modulation effect of these factors can be eliminated by the design of spatial non-uniform illumination. Therefore, a new single pixel imaging method is proposed in this paper. In order to enhance the image contrast and improve the image quality, the method reconstructs the spectrum of the imaging object by projecting two-dimensional cosine gray fringes with spatial modulated amplitude. The surface reflectance of the object is imaged, and the experimental results are consistent with the theoretical analysis, which verifies the effectiveness and feasibility of the method.

To improve the imaging quality of ghost imaging systems in turbid media, we propose an underwater polarization difference ghost imaging method based on histogram preprocessing. Firstly, two images in orthogonal polarization directions are acquired by polarization detection. Then, the images are preprocessed by gray-scale stretching. Finally, the reconstructed image is obtained with the polarization difference algorithm. Experimental results show that proposed method can not only improve the image quality degraded by scattering in turbid media but also enhance image details. In addition, compared with the existing polarization ghost imaging method, the gray-scale statistical analysis using the gray-scale histograms of the reconstructed images shows that when the medium has a higher turbidity concentration, the proposed method can still distinguish the target from the turbid medium.

To solve the problem that the shearing amount in traditional spatial carrier phase-shift shearography systems is hard to be rapidly, quantitatively, and precisely modulated, this paper proposes a spatial carrier phase-shift shearography system based on a liquid crystal spatial light modulator (LC-SLM). With LC-SLM as a shearing amount modulator, the phase of each tiny liquid crystal unit in the working area of LC-SLM is precisely controlled by electrical addressing, and the phase-shift modulation in the frequency domain is converted into displacement modulation in the spatial domain through the Fourier transform of lens. As a result, the shearing amount can be modulated accurately. An experimental setup is developed to verify the proposed method. The experimental results show that this system can carry out the full-field measurement of interference phase changes caused by object deformation. By writing the periodic fringe images in LC-SLM, this system can modulate the shearing amount and the spatial carrier frequency. Compared with the traditional spatial carrier phase-shift shearography system, the proposed system can quantitatively, accurately, and rapidly modulate the shearing amount.

Image noise will seriously affect the measurement accuracy of the digital volume correlation (DVC) method. Therefore, this paper studies the effect of image noise and Gaussian prefiltering on the DVC displacements measurement results through numerical simulation translation experiments, real rescan and compression experiments. The experimental results show that, Gaussian prefiltering can significantly reduce the noise level of the images, and thus mitigates the noise-induced bias error, but has little effect on the interpolation error. The reason is that the Gaussian prefiltering will reduce the gray gradient of the image. Therefore, Gaussian prefiltering can significantly improve the accuracy of the classic forward additive Newton-Raphson algorithm. But it has little effect on the accuracy of the inverse compositional Gauss-Newton algorithm. Further, the Gaussian prefiltering is used to reduce image noise, which can improve the correlation coefficient of each calculation point in the 2 algorithms, and obtain a sub-volume with a higher degree of matching.

Multi-beam laser is one of the hot topics in aerospace topographic mapping, where the core is the determination of laser pointing. Laser reference cameras (LRCs) are usually used for recording, and the boresight monitoring function and performance of LRCs are the key to ensuring the accuracy of laser pointing determination. Firstly, a method based on a high-stability central prism is proposed to monitor and measure the LRC boresight, and the calculation algorithm is given. Second, an experimental verification system is built, and the correctness of the algorithm is verified with a Hexpod turntable and a photoelectric auto-collimator. The results show that the measurement results obtained by the proposed algorithm are consistent with those obtained by the photoelectric auto-collimator, with the error of less than 0.2″. Furthermore, the thermal drift of the LRC boresight is measured, and it reaches 0.3″ and 1.4″ around the X axis and Y axis, respectively. Finally, the monitoring accuracies of the boresight in different environments are evaluated. The results show that the monitoring accuracy in vacuum environment is 0.02″(one times standard deviation), which meets the practical application requirements of the on-orbit monitoring of the LRC boresight. Compared with the vacuum condition, the error increases 2--4 times under the atmospheric vibration isolation condition and 10--15 times under the non-vibration isolation condition, respectively.

In light of the situation that it is difficult to accurately detect pedestrians in real scenes due to mutual occlusion, a feature extraction enhanced detection algorithm based on attention mechanism is proposed. Firstly, attention modules are added to learn the relationship between feature channels and the spatial information of feature maps, so as to enhance feature extraction in the visual area of pedestrian targets. Secondly, according to the actual size of pedestrian data, the k-means++ algorithm is used to cluster pedestrian labels, so as to determine the size and proportion of anchors. Distance-intersection over union loss function (DIOULoss) is used to design the loss function of the detector, so that the regression of the detection box pays more attention to the intersection over union between the candidate box and the real box, as well as the center distance between the two boxes. Finally, a new non-maximum suppression algorithm (DSoft-NMS) is presented to preserve more accurate prediction boxes. The proposed method has been tested on CityPersons and WiderPerson datasets, and the results show that the proposed method with a simple network structure has higher detection accuracy in occluded pedestrian detection, which is convenient for subsequent research.

In this paper, regarding the phenomenon that the basic feature point matching algorithm is prone to mismatch in visual odometry, we proposes a ring matching algorithm of feature points combined with the bidirectional optical flow method. This algorithm forms a ring structure between the stereo image and the images in the front and rear frames. For the images in the front and rear frames, the bidirectional pyramid optical flow method is used to track feature points and eliminate mismatched feature points. The basic feature point matching algorithm usually adopts fast library for approximate nearest neighbors (FLANN), but the result contains many mismatched point pairs. The proposed matching algorithm can not only eliminate the mismatched feature points but also make the feature points evenly distributed on the images. Subsequently, the perspective-3-point (P3P) algorithm based on 3D-2D points is combined with random sample consensus (RANSAC) to obtain the initial pose estimation results. The general graph optimization (g2o) library is employed to further optimize the pose estimation results. The positioning experiments verify that the ring matching algorithm of feature points combined with the bidirectional optical flow method has higher positioning accuracy.

In the spectral domain optical coherence tomography system, the linear-wavenumber spectrometer can not only reduce the interpolation error of the image and improve the image quality, but also improve the imaging speed and sensitivity of the system. Therefore, in view of the shortcomings of the existing grism spectroscopic optical system, a method combining grism structure spectroscopy and object image aberration compensation is proposed in this paper. By introducing distortion and vertical axis chromatic aberration, extremely high wavenumber linearity is achieved. A linear-wavenumber spectrometer with a working band of 750--950 nm, a spectral resolution of 0.1 nm, and a length of the system is about 148 mm is designed. The design results show that the wavenumber linearity of the spectrometer reaches 0.0056, which is about 11 times that before aberration compensation. Simulation analysis results show that the spectroscopic system can significantly improve the sensitivity of the optical coherence tomography system. Compared with the uncompensated system, the sensitivity of the compensated system at an optical path difference of 3.5 mm is increased by 7.8 dB.

How to make full use of the full-area radiant energy of parabolic trough collector (PTC) under a large opening area is a key issue that needs to be considered. For this reason, a new trough free-form solar concentrato (TFFC) is proposed, which uses the form of photovoltaic/photothermal simultaneous collection to expand the PTC concentrator. The opening utilization area of the optical device adopts the traditional trough heat exchanger form on the inner side close to the PTC, and the outer side is based on the edge light transmission principle to design a free-form surface concentrating photovoltaic system, which uses an inclined method to expand the energy receiving area of the solar panel, and obtains a very uniform concentrating energy flow distribution. After that, the line tracing method is used to verify its optical characteristics, and compared with the traditional PTC system to complete the sensitivity analysis of structural parameters and the discussion of error factors. The results obtained are of great significance to the improvement and optimization of the trough solar collector system.

The existing radiation cooling device is not self-adaptive, that is, it can only lower the temperature but not increase the temperature. In order to achieve the dual functions of heating and cooling, and can automatically switch with the change of ambient temperature, this article focuses on an adaptive temperature regulator based on vanadium dioxide, and systematically discusses its photothermal control principles, technical bottlenecks, and implementation methods. The combination method of transmission matrix and genetic algorithm is used to design and optimize the multilayer structure integrated by the resonant cavity and the filter, and obtain the excellent photothermal control performance of solar absorption ratio of 0.1, high temperature emissivity of 0.76 and emissivity difference of 0.7. On this basis, the real-time temperature response of the optimized structure is predicted according to the actual environment, and its adaptive temperature control function is verified theoretically. This research provides an alternative solution for adaptive temperature control based on vanadium dioxide, and provides a corresponding technical reserve for the development of intelligent thermal control technology.

Photoelastic modulator (PEM) is a kind of light modulator widely used in optical detection, which uses inverse piezoelectric effect of piezoelectric crystal to periodically change refractive index inside photoelastic crystal to realize phase modulation of optical signal passing through photoelastic crystal. PEM has the highest modulation efficiency when the system is in a resonant state. Therefore, resonant frequency and quality factor representing its working efficiency, are two important parameters of PEM. To study the resonance characteristics of PEM, we design a two-dimensional octagonal symmetric structure PEM with target frequency near 50 kHz and verify it theoretically and experimentally. At the same time, we propose a PEM resonance characteristic measurement method based on impedance analysis. In this paper, the theoretical parameters of the target PEM are obtained by establishing a frequency model for analysis, and then theoretical verification is carried out by numerical simulation software. Finally, two groups of PEM samples are prepared. In addition, the resonance characteristics of the samples are experimentally verified based on the impedance analysis method. The measured resonant frequencies of the sample PEMs are 52.363 kHz and 52.353 kHz, and the quality factors are 5071.2 and 6096.7, respectively.

Firstly, according to intensity distributions of reflected light and transmitted light in perpendicular and parallel directions on surface of transparent objects, the functional relation between the components of reflected light and transmitted light and that between the polarization degrees of reflected light and transmitted light are derived with the principle of polarization orthogonal decomposition in this paper. Then, based on the imaging principle of camera, the angles between incident lights and normal vectors and the azimuth angles of the incident planes at different positions of transparent object surface are obtained by using surface normal vector to summary the distribution rules of the polarization degrees of reflected light and transmitted light at each pixel in the image. Finally, according to the polarization characteristics and correlation features between reflection and transmission, the minimum of normalized cross-correlation between reflection component and transmission component is calculated by gradient descent algorithm to realize the separation of reflected light and transmitted light, which suppresses the interference of reflected light on the surface of transparent objects, and it plays an important role in information processing and applications such as target detection and image matching in environment with complex reflected light.

When a laser is disturbed by low-frequency factors (e.g., thermal fluctuations and mechanical vibrations), laser wavelength drift and light intensity fluctuations occur, which affect the measurement accuracy of the system. In this study, a 2.73-μm distributed feedback (DFB) laser is used as the detection light source and a set of CO2 gas-sensing system based on hollow waveguide fiber is built. The proposed system uses the third harmonic absorption signal (3f) of CO2 to achieve the frequency stabilization of the laser and then uses the harmonic signal to reverse the concentration of CO2. In the experiment, the standard CO2 gas is measured for a long time and the system is analyzed according to four measurement methods: second harmonic (2f), stabilized second harmonic (2f-lock), second harmonic ratio first harmonic (2f/1f), and stabilized second harmonic ratio first harmonic (2f/1f-lock). The system measurement accuracy obtained under the 2f-lock condition is 0.001255, which is 2.4 times higher than that under the 2f condition. The system measurement accuracy obtained under 2f/1f-lock condition is 0.00138, which is 2.34 times higher than that under 2f/1f condition. It can be seen that the 2f/1f-lock method has the longest stable time of 210 s and the lowest detection limit of 4.24×10 -5. Comprehensive analyses reveal that 2f/1f-lock is the most optimal among the four methods. Kalman filtering is performed on the measured data under the 2f/1f-lock condition, and the measurement accuracy of the filtered system is 0.0002786, which is 4.95 times higher than that before the filtering.

With the development of optoelectronic technologies, low-light imaging and its application under low illumination conditions have become one of the research hotspots in recent years. For the meteorological and marine environment satellites of quantitative remote sensing, the research on radiometric calibration and data application of low-light imagers has also received widespread attention. This paper takes the most representative low-light imager load VIIRS/DNB in the world as an example, and systematically illustrates the research progress on the sensor absolute radiation calibration and data application in the meteorological and marine environment under different moon phases at night, in order to provide useful information for scientific research and practical engineering applications.

Based on the random frog algorithm, a window-based random frog algorithm is proposed. With a continuous window instead of a single wavelength point, the proposed algorithm improves the optimization accuracy of the original random frog algorithm and reduces the iteration times of the algorithm, thus improving the convergence speed. The results of the blood samples show that compared with the full-band results, the root mean square error of prediction (RMSEP) of the built partial least square model decreases by 47.9%, and the correlation coefficient of the prediction set, Rp, increases by 4.07%, proving the validity of the proposed algorithm. The regression analysis of the characteristic wavelengths selected by the three mainstream algorithms and the window-based random frog algorithm is carried out, which demonstrates the superiority of the improved algorithm in selecting the characteristic wavelength.

The short-wave ultraviolet (UVC) Raman spectrometer has the advantages of high Raman scattering intensity, weak fluorescence noise, and low background light noise. In order to realize the miniaturization of the UVC Raman spectrometer and extend its application fields, the key technologies of all solid-state 266 nm laser Raman spectrometer are studied. The diode pumped 266 nm solid-state laser is used as the excitation light source to effectively reduce the system volume and improve the system robustness. A solid-state spectrometer with Littrow structure is designed for obtaining the UVC Raman spectral signals, in which a single off-axis parabolic reflector is used as a collimating/focusing lens, and a plane grating with high line density is used as a dispersion element. The resolution in the band of 269--293.5 nm is better than 0.07 nm using a 50 μm slit, reaching the Raman spectral resolution requirement of 10 cm -1. A complete Raman spectrometer is built to test the Raman characteristic peak of ethyl alcohol with 99.8% mass fraction, and the feasibility and rationality of the above system is verified.

Micro-electro-mechanical system (MEMS) infrared conversion film, which can convert write visible light images into infrared images, is widely used in infrared target simulators. When an infrared focal plane detector is docking with an infrared target simulator, the integration time of detector must be greater than the display time of infrared image to collect all the gray scales of infrared image. A light-driven technology based on MEMS infrared conversion film is proposed in this paper. Energy of write visible light image is integrated with thermal inertia of the film, which reduces requirement for integration time of infrared detector. Temperature characteristics of the film in the light-driven state are simulated and tested. Results show that the gray value of infrared image generated by the film is linearly related to that of write visible light image. The infrared image has heating time of 9 ms, keeping time of 1 ms, and cooling time of 10 ms. The detector can get a 256 gray-scale infrared image by integrating for any period during keeping time. The maximum temperature rise of infrared image is 112 K, and the temperature resolution corresponding to the unit gray value of infrared image is 0.44 K.

Based on the Bragg grating/metallic film asymmetric resonant cavity, a multi-layer film structure that enhances the ultraviolet absorption of graphene is designed. The transfer matrix method is used to simulate and calculate the correlation between the film structure parameters and the absorption spectra of graphene. As a result, the optimal structure parameters are obtained. The calculation results show that the resonance effect of the asymmetric resonant cavity strongly promotes the interaction between incident light and graphene. After optimization, the ultraviolet absorption of graphene at 275 nm is up to 0.9534 and the full width at half maximum is only 2.0 nm. Further research indicates that the central wavelength of the Bragg grating and the incident angle of light can regulate the wavelength and size of the graphene absorption peak. Increasing the number of graphene layers is able to enhance the ultraviolet absorption of graphene and expand absorption spectra. The proposed structure can also promote the ultraviolet absorption of other wide-band-gap two-dimensional materials.

An online monitoring device is designed for the real-time monitoring of hard X-ray beams at Shanghai Synchrotron Radiation Facility (SSRF). Compared with the existing beam position monitors and gas ionization chambers, it can monitor the beam position, size, and flux, and reconstruct the relative phase variation of the incident wavefront with an acquisition speed of 2 kHz, without affecting the downstream experiments. In addition, it has potential applicability for feedbacks of the optical devices and experimental positions in super-long and micro-nano-focused beamline stations. Experiments are carried out in the USAXS and micro-focusing SAXS experimental stations at SSRF, and beam stability data are collected at common experimental frame rates. The time-frequency domain signals of position, flux, and size distribution of the incident beam are analyzed, and the phase variation of the incident wavefront is obtained as well. Finally, the feasibility of this device is verified.

In combination with ferromagnetic resonance and magnetic circular dichroism, the time-resolved X-ray ferromagnetic resonance method based on synchrotron radiation devices, which adopts the “pump-probe” technique of phase shift mode, is a unique technique to study a wide class of spintronics issues such as spin transfer torque and spin current. In this paper, an experimental apparatus that allows for the combination of the picosecond time-resolved technique and X-ray ferromagnetic resonance is introduced for the first time at the third-generation synchrotron radiation light source in China, Shanghai Synchrotron Radiation Facility. With this apparatus, the single-layer Permalloy (Ni81Fe19, Py) undergoes ferromagnetic resonance under the 2.5 GHz microwave excitation and we measure the time-dependent projection of electronic spin precession cone angle of Ni in the Ni81Fe19 film in the beam direction. The results show that the apparatus can excite the electronic spin precession of magnetic elements on the magnitude of 2.5 GHz and detect the amplitude and phase of electronic spin precession on the picosecond scale. This apparatus can provide unique technical support for the investigation into the spintronic materials and devices.