A crosstalk information correction method for a universal mosaic multispectral imaging crop growth sensor device

Abstract

The present invention discloses a universal mosaic multi-spectral imaging crop growth sensor device crosstalk information correction method, including: S1 using a sensor device combined with an adjustable monochromatic light source system to step-by-step obtain uniform light source images of different bands; S2 searching for uniform light source images of corresponding bands according to the band settings on the mosaic filter, traversing and recording the pixel index with the largest response value in the macro-pixel area of ​​each band image. The response value of the pixel position corresponding to each step band image is extracted according to the pixel index, and the response average value curve of each channel is drawn; S3 draws a Gaussian response curve according to the response value curve of each channel, and uses the pseudo-inverse matrix of the response value data to multiply the Gaussian data matrix to obtain a correction coefficient matrix, and uses the matrix to eliminate data crosstalk between the original spectral image bands. The present invention solves the problem of band crosstalk information of mosaic multi-spectral imaging crop growth sensor devices, and has the characteristics of universality and high efficiency.

Description

Technical Field

[0001] The present invention relates to the field of smart agriculture, and in particular to a method for correcting crosstalk information of a universal mosaic multispectral imaging crop growth sensor device. Background Art

[0002] Accurate perception of crop growth information is a key technical link in smart agriculture, which can promote the quantification and intelligence of agricultural production management. Spectral imaging technology, with its technical advantage of integrating graphs and spectra, provides a technical means for real-time, accurate and non-destructive acquisition of crop growth information, while multispectral imaging sensors provide a reliable basic tool for the realization of spectral imaging technology. Current multispectral imaging crop growth sensors usually use multiple sets of narrow-band interference filters for spectroscopic imaging, and their assembly structure is relatively simple. However, with the increase in bands, the optomechanical structure of such equipment becomes more complex, which increases the difficulty of image registration and fusion between different bands, and also limits the real-time processing capability of graph information, thus affecting the real-time and efficiency of crop growth monitoring.

[0003] Mosaic spectral imaging spectrometry technology can coat multiple bands on the same chip-level coated filter, and obtain crop map information of multiple bands in a snapshot manner through a single exposure. This method not only promotes the real-time processing of crop map information, but also promotes the miniaturization and integration of sensing equipment. However, since the mosaic filter needs to be installed on the surface of the detector, there is inevitably a certain integration spacing between the mosaic filter and the detector, which causes different band information to enter the adjacent or other channel areas, thereby causing information crosstalk between different bands. Therefore, correcting the crosstalk information by a reasonable method is crucial to ensure the accuracy of spectral data. The patent (application number CN202310860144.3) intends to disclose a snapshot-mosaic multi-spectral imaging crop growth sensor device crosstalk information removal method. However, this method only performs downsampling when the mosaic filter is perfectly aligned with each channel of the detector, and does not consider other complex situations. In addition, this method only uses a simple least squares method to obtain the crosstalk correction coefficient matrix, and does not fully consider the case where the response value matrix is ​​a non-square matrix. Therefore, there is an urgent need for a universal mosaic multispectral imaging crop growth sensor device crosstalk information correction method to solve the information crosstalk problem between different bands, thereby improving the accuracy of spectral data and ensuring the accuracy and efficiency of crop growth monitoring. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the above-mentioned background technology and to provide a universal mosaic multispectral imaging crop growth sensor device crosstalk information correction method, which can correct the crosstalk information between different bands in real time and accurately.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is: S1 uses a mosaic multi-spectral imaging crop growth sensor device combined with an adjustable monochromatic light source system to step-by-step obtain uniform light source images of different bands; S2 searches for uniform light source images of corresponding bands according to the band settings on the mosaic filter, traverses and records the pixel index with the largest response value in the macro-pixel area of ​​each band image. The position of each band in the mosaic filter is determined by these indexes. According to the pixel index, the response value of the pixel position corresponding to the uniform light source image of each band after the adjustable monochromatic light source system is step-by-step split is extracted, and the response value curve of each channel is drawn; S3 draws the corresponding Gaussian response curve according to the response value curve of each channel, and uses the pseudo-inverse matrix of the response value data to multiply the Gaussian data matrix to obtain the correction coefficient matrix, and uses the matrix to eliminate the data crosstalk between the original spectral image bands. The present invention effectively solves the problem of spectral information crosstalk caused by the integrated spacing between the mosaic filter and the detector, and has the characteristics of generalization and simplification.

[0006] Further, the S1 includes:

[0007] S11 fixes the mosaic multispectral imaging crop growth sensor device on the light outlet end of the integrating sphere of the adjustable monochromatic light source system, aligns and parallels the imaging objective lens of the sensor device with the center hole of the integrating sphere, and ensures that the field of view of the imaging objective lens completely covers the internal area of ​​the integrating sphere.

[0008] S12 connects the computer to the mosaic multi-spectral imaging crop growth sensor device and the adjustable monochromatic light source system, sets the exposure time and gain value of the mosaic multi-spectral imaging crop growth sensor device to 100ms and 6dB, adjusts the adjustable monochromatic light source system according to the band of the mosaic filter, and finds the band with the highest grayscale value in the mosaic filter; under the condition of the band with the highest grayscale value, adjusts the grayscale value of the mosaic multi-spectral imaging crop growth sensor device to 800-1000, and sets the exposure time and gain value at this time as the reference value;

[0009] S13 uses a computer to adjust the band of the adjustable monochromatic light source system. First, the minimum band value of the adjustable monochromatic light source system is set to be smaller than the minimum band value of the mosaic filter, and the maximum band value is set to be larger than the maximum band value of the mosaic filter. Then, the band of the adjustable monochromatic light source system is adjusted step by step at an interval of 2 nm between the minimum band and the maximum band.

[0010] S14 uses a computer to control the mosaic multi-spectral imaging crop growth sensor device to collect uniform light source images of each band that are transmitted to the integrating sphere after the step-by-step splitting of the adjustable monochromatic light source system band.

[0011] Further, the S2 includes:

[0012] S21 searches for the corresponding band images in the collected uniform light source images of different bands according to the band settings on the mosaic filter; determines the macro pixel area and its distribution law of each channel according to the size of each channel of the mosaic filter in the mosaic multispectral imaging crop growth sensor device; traverses and records the pixel index with the highest response value in all macro pixel areas to determine the pixel position corresponding to each band on the mosaic filter;

[0013] S22 extracts the response value of the position in all step images according to the determined pixel position of each band, calculates the average response value on each step image, draws the response value curve of each channel, and combines these response value data into a matrix D, that is, the original matrix.

[0014] Further, the S3 includes:

[0015] S31 draws the Gaussian curve corresponding to each channel according to the original response value curve of each channel in step S22, and records the Gaussian data matrix of each channel as D ideal , which is the target matrix;

[0016] S32 converts the pseudo inverse matrix D of the original matrix D in step S22 into + and the target matrix D in step S31 ideal Multiply them together to obtain the crosstalk correction coefficient matrix T; minimize the Euclidean norm of the error between the target value and the original value, that is:

[0017]

[0018] Among them, X and X ideal represent the actual measured and ideal response data matrices respectively;

[0019] S33 All the original spectral image crosstalk information is corrected by the crosstalk correction coefficient matrix T in step S32, and the correction formula used is as follows:

[0020]

[0021] Among them, P i represents the reflectance value matrix after correction of the i-th band, T ij represents the correction coefficient at the jth row and ith column in the correction coefficient matrix T, PX j It represents the original reflectance value matrix of the jth band of the original spectral image to be corrected, and n represents the number of bands.

[0022] Furthermore, the mosaic multispectral imaging crop growth sensor device includes: an imaging objective lens, a mosaic multispectral camera 1, a mosaic multispectral camera 2, a front fixing device, a main control system fixing part, a main control system and a cover, wherein the main control system fixing part fixes the main control system at the rear end of the cover; the front fixing device fixes the imaging objective lens, the mosaic multispectral camera 1, and the mosaic multispectral camera 2 at the front end of the cover in sequence, and the light passes through the imaging objective lens and is received by the mosaic multispectral camera 1 and the mosaic multispectral camera 2 in sequence, and the mosaic multispectral camera 1 and the mosaic multispectral camera 2 are connected to the main control system to send data thereto. The main control system performs real-time online processing and interpretation of the received atlas information.

[0023] The beneficial effects of the present invention are as follows: uniform light source images in different bands are obtained through a mosaic multispectral imaging crop growth sensor device, thereby ensuring the uniformity of illumination of images in different bands; the position of each channel is determined according to the pixel response characteristics and macropixel distribution law in the macropixel, the image downsampling method is clarified, and the accuracy of image response value extraction of each channel is ensured; the response data of each channel is used as the original matrix, and the Gaussian curve data is used as the target matrix, and the crosstalk correction coefficient matrix is ​​obtained by multiplying the pseudo-inverse matrix of the original matrix with the target matrix, thereby ensuring that the Euclidean norm of the error between the target value and the original value is minimized; the real-time online correction of the crosstalk information of the original image is realized through the crosstalk correction coefficient matrix, thereby effectively solving the interference problem of the crosstalk information between different bands on the accuracy of the spectral data. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a schematic diagram of the structure of the mosaic multi-spectral imaging crop growth sensor device in the present invention.

[0025] FIG. 2 is a schematic diagram of the corresponding wavelength bands of the mosaic filter of the mosaic multispectral imaging crop growth sensor device of the present invention.

[0026] Figure 3 Schematic diagram of the process of removing spectral crosstalk information of the mosaic multi-spectral imaging crop growth sensor device in the present invention

[0027] Figure 4 This is a correlation diagram between the data matrix after spectral crosstalk information correction of the mosaic multi-spectral imaging crop growth sensor device of the present invention and the Gaussian data matrix

[0028] Figure 5 Statistical graph of the reflectivity of the diffuse reflector before and after the spectral crosstalk information of the mosaic multi-spectral imaging crop growth sensor device of the present invention is removed

[0029] Marked in the figure: 1-imaging objective lens, 11-imaging objective lens 1, 12-imaging objective lens 2, 2-mosaic multispectral camera 1, 21-protective glass 1, 22-mosaic filter 1, 23-CMOS image sensor 1, 3-mosaic multispectral camera 2, 31-protective glass 2, 32-mosaic filter 2, 33-CMOS image sensor 2, 4-front fixing device, 5-main control system fixing part, 6-main control system, 7-cover. DETAILED DESCRIPTION

[0030] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] like Figure 1 As shown in FIG2 , the mosaic multispectral imaging crop growth sensor device includes an imaging lens 1, a mosaic multispectral camera 1 2, a mosaic multispectral camera 2 3, a front fixing device 4, a main control system fixing member 5, a main control system 6 and a cover 7 (in a preferred embodiment, the imaging lens 1 uses a V2520-MPZ lens from Computar Optics of Japan; the main control system 6 is powered by NVIDIA Jetson The TX2 embedded AI module is built as a prior art; the mosaic multi-spectral camera refers to the applicant's prior application 2024101689797. Compared with the mosaic multi-spectral camera of the prior application, this example adds a group of mosaic multi-spectral cameras with other central bands to enhance the reliability of the method proposed in this patent. The mosaic filter 1 22 and the mosaic filter 2 32 are respectively arranged on the surface of the CMOS image sensor 1 23 and the CMOS image sensor 2 32. Next, the specific implementation method is described by taking the mosaic multi-spectral camera 1 2 with bands 458, 487, 527 and 558nm as the central bands and the mosaic multi-spectral camera 2 3 with bands 644, 716, 737 and 813nm as the central bands as examples.

[0032] In the present invention, in order to ensure that the images of each band acquired by the mosaic multi-spectral imaging crop growth sensor device are uniform light source images, an adjustable monochromatic light source system of Beijing Zhuoli Hanguan Instrument Co., Ltd. is adopted. The system contains a wide-spectrum light source with a spectral coverage range of 350-2500nm, a spectrometer with a spectral range of 350-1100nm, an inner diameter and a light output aperture of 200 and 50mm respectively, and an integrating sphere with a uniformity of >98%. The mosaic multi-spectral imaging crop growth sensor device is used to acquire uniform spectral images of different bands.

[0033] like Figure 3 As shown, a universal mosaic multispectral imaging crop growth sensor device crosstalk information correction method includes the following process steps:

[0034] Step S1: Use a computer to control the mosaic multispectral imaging crop growth sensor device to obtain uniform light source images of different bands; Step S1 is specifically divided into the following three small steps:

[0035] S11 fixes the mosaic multispectral imaging crop growth sensor device on the light outlet end of the integrating sphere of the adjustable monochromatic light source system, aligns and parallels the imaging objective lens of the sensor device with the center hole of the integrating sphere, and ensures that the field of view of the imaging objective lens completely covers the internal area of ​​the integrating sphere.

[0036] Step S12: Connect the computer to the mosaic multi-spectral imaging crop growth sensor device and the adjustable monochromatic light source system. First, adjust the exposure time and gain value of the mosaic multi-spectral camera 2 to 100ms and 6dB respectively through the computer, and adjust the band of the spectrometer of the adjustable monochromatic light source system to 458, 487, 527 and 558nm respectively, wherein the image grayscale value is the maximum value under the 558nm band setting, and then adjust the grayscale value of the mosaic multi-spectral camera 2 to 880-920 at 558nm, and use the exposure time of 950ms and the gain value of 4dB at this time as the benchmark values ​​of the mosaic multi-spectral camera 2. Then the bands of the adjustable monochromatic light source system spectrometer are adjusted to 644, 716, 737 and 813nm respectively, where the image grayscale value is the maximum under the 813nm band setting. Then, at 813nm, the grayscale value of the mosaic multispectral camera 23 is adjusted to 880-920, and the exposure time of 850ms and the gain value of 2.5dB at this time are used as the benchmark values ​​of the mosaic multispectral camera 23.

[0037] Step S13: Align the imaging objective lens 11 of the mosaic multi-spectral camera 2 with the inside of the integrating sphere, use a computer to adjust the band of the adjustable monochromatic light source system spectrometer to 350nm, and step the band value of the adjustable monochromatic light source system spectrometer at intervals of 2nm. The step adjustment range is 350-700nm. During each step adjustment, the mosaic multi-spectral camera 2 is controlled to collect images of each band; align the imaging objective lens 2 12 of the mosaic multi-spectral camera 2 3 with the inside of the integrating sphere, use a computer to adjust the band of the adjustable monochromatic light source system spectrometer to 500nm, and step the band value of the adjustable monochromatic light source system spectrometer at intervals of 2nm. The step adjustment range is 500-900nm. During each step adjustment, the mosaic multi-spectral camera 2 3 is controlled to collect images of each band.

[0038] Step S2: According to the band setting on the mosaic filter, find the uniform light source image of the corresponding band, traverse and record the pixel index with the largest response value in the macro pixel area of ​​each band image. Extract the response value of the pixel position corresponding to each step band image according to the pixel index, and draw the response average curve of each channel; Step S2 is specifically divided into the following two small steps:

[0039] S21 searches for the corresponding band images in the acquired step image according to the bands (458, 487, 527, 558, 644, 716, 737 and 813 nm) of the mosaic multispectral camera 1 2 and mosaic multispectral camera 2 3 (as shown in FIG2 ). According to the size of each channel of the mosaic filter 1 22 and mosaic filter 2 32 (11×11 μm), the macro pixel area of ​​each channel is determined. The macro pixel area is distributed over the entire imaging target surface at an interval of 2 pixels (such as Figure 3 a). MATLAB software is used to traverse and record the pixel index with the highest response value in all macro-pixel areas to determine the pixel position corresponding to each band on mosaic multispectral camera 1 2 and mosaic multispectral camera 2 3;

[0040] S22 uses MATLAB software to extract the response value of each stepping image according to the pixel position of each band determined in step S21, calculates the response average value of each image, and draws the response average value curve of each channel (such as Figure 3 b), and these response value data are combined into a matrix D, which is the original matrix.

[0041] Step S3: Draw a Gaussian response curve according to the response value curve of each channel, and use the pseudo-inverse matrix of the response value data to multiply the Gaussian data matrix to obtain the correction coefficient matrix, and use the matrix to eliminate the data crosstalk between the original spectral image bands; Step S3 is specifically divided into the following 4 small steps:

[0042] S31 uses MATLAB software to draw the Gaussian curve corresponding to each channel according to the original response value curve of each channel in step S22 (such as Figure 3 c), and the Gaussian data matrix of each channel is recorded as D ideal , which is the target matrix;

[0043] S32 converts the pseudo inverse matrix D of the original matrix D in step S22 into + and the target matrix D in step S31 ideal Multiply and solve the crosstalk correction coefficient matrix T (such as Figure 3 d), so that the Euclidean norm of the error between the target value and the original value is minimized, that is:

[0044]

[0045] Among them, X and X ideal Represent the actual measured (including crosstalk) and ideal response data matrices respectively.

[0046] S33 uses the crosstalk correction coefficient matrix T to correct the original response value matrix D, and the response curves of each channel are as follows: Figure 3 (e) as shown.

[0047] In the present invention, the proposed crosstalk information correction method for a universal mosaic multispectral imaging crop growth sensor device is verified by two methods: verification method 1 uses the Gaussian curve constructed in step S31 (such as Figure 3 c) data matrix and the channel curves after crosstalk correction obtained in step S33 (as shown in Figure 3 e) is achieved by the data matrix correlation results.

[0048] In the present invention, Figure 4 As shown in Figure 2, the implementation effect of the proposed crosstalk information correction method for a universal mosaic multispectral imaging crop growth sensor device is as follows: Figure 4 . Figure 4 The Pearson correlation analysis results of the Gaussian curve data matrix of each channel and the corrected data matrix are shown in the figure. The correlation coefficients of the eight channels are 0.979, 0.995, 0.995, 0.989, 0.994, 0.997, 0.998 and 0.983 respectively. The results show that the data matrix corrected by the crosstalk correction coefficient matrix has a very good correlation with the target matrix, proving that the proposed crosstalk information correction method has a very good effect.

[0049] Verification method 2 uses a green diffuse reflectance calibration plate with a reflectivity of 20%-30% from Guangzhou Changhui Electronic Technology Co., Ltd. The digital quantization value of the image obtained by the mosaic multispectral imaging crop growth sensor device is converted into a reflectance value through radiation correction, and the standard calibration cloth used is a gray calibration cloth with a reflectivity of 40%. The calculation formula used is as follows:

[0050]

[0051] Among them, DN c Indicates the pixel response value of the green diffuse reflection calibration plate to be tested, DN w Represents the pixel response value of the gray calibration cloth obtained, DN b represents the dark field pixel response value (obtained by taking an image with the lens cap covered), ρ w Indicates that the reflectance value of the gray calibration cloth is 0.4.

[0052] S34 uses the crosstalk correction coefficient matrix T to eliminate the original digital quantization value of the green diffuse reflection calibration plate obtained, and the correction formula used is as follows:

[0053]

[0054] Among them, P i represents the reflectance value matrix after correction of the i-th band, T jirepresents the correction coefficient at the jth row and ith column in the correction coefficient matrix T, PX j It represents the original reflectance value matrix of the jth band of the original spectral image to be corrected, and n represents the number of bands.

[0055] In the present invention, the implementation effect of the crosstalk information correction method of a universal mosaic multispectral imaging crop growth sensor device is shown in Figure 5 . Figure 5 The figure shows the effect of the reflectance value of the green diffuse reflection calibration plate collected by the mosaic multispectral imaging crop growth sensor device before and after correction. Figure 5 As shown in the figure, before the data correction using the crosstalk correction coefficient matrix T, the reflectance value of the measured green diffuse reflectance calibration plate is quite different from the actual value, with the maximum and average relative errors being 26.49% and 10.07% respectively. After the data correction using the crosstalk correction coefficient matrix T, the crosstalk information between the bands has a good correction effect, and the maximum and average relative errors are reduced to 4.85% and 1.74% respectively. This result shows that the proposed crosstalk information correction method for the universal mosaic multispectral imaging crop growth sensor device can be fully used in the measured data, thereby improving the accuracy of crop growth monitoring.

[0056] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the protection scope of the present invention in any form, and all technical solutions obtained by equivalent replacement and the like fall within the protection scope of the present invention.

[0057] The parts not involved in the present invention are the same as the prior art or can be implemented by using the prior art.

Claims

1. A universal mosaic multispectral imaging crop growth sensor device crosstalk information correction method, characterized in that: The following steps are involved: S1 uses a mosaic multispectral imaging crop growth sensor device combined with an adjustable monochromatic light source system to step-by-step acquire uniform light source images of different bands; S2 searches for a uniform light source image of a corresponding band according to the band setting on the mosaic filter, traverses and records the pixel index with the largest response value in the macro pixel area of ​​each band image; determines the pixel position corresponding to each band in the mosaic filter through these indexes; extracts the response value of the pixel position corresponding to the uniform light source image of each band after the adjustable monochromatic light source system is stepped and split according to the pixel index, and draws the response average value curve of each channel; S2 includes: S21 searches for the corresponding band images in the collected uniform light source images of different bands according to the band settings on the mosaic filter; determines the macro pixel area and its distribution law of each channel according to the size of each channel of the mosaic filter in the mosaic multispectral imaging crop growth sensor device; traverses and records the pixel index with the highest response value in all macro pixel areas to determine the pixel position corresponding to each band on the mosaic filter; S22 extracts the response value of the position in all step images according to the determined pixel position of each band, calculates the average response value on each step image, draws the response value curve of each channel, and forms the matrix D, i.e., the original matrix, with these response value data; S3 draws the corresponding Gaussian response curve according to the response value curve of each channel, and multiplies the pseudo-inverse matrix of the response value data with the Gaussian data matrix to obtain the crosstalk correction coefficient matrix, and uses the matrix to eliminate the data crosstalk between the original spectral image bands; S3 includes: S31 draws the Gaussian curve corresponding to each channel according to the original response value curve of each channel in step S22, and records the Gaussian data matrix of each channel as D ideal , which is the target matrix; S32 converts the pseudo inverse matrix D of the original matrix D in step S22 into + and the target matrix D in step S31 ideal Multiply them together to obtain the crosstalk correction coefficient matrix T; minimize the Euclidean norm of the error between the target value and the original value, that is: Among them, X and X ideal represent the actual measured and ideal response data matrices respectively; S33 All the original spectral image crosstalk information is corrected by the crosstalk correction coefficient matrix T in step S32, and the correction formula used is as follows: Among them, P i represents the reflectance value matrix after correction of the i-th band, T ij represents the correction coefficient at the jth row and ith column in the correction coefficient matrix T, PX j It represents the original reflectance value matrix of the jth band of the original spectral image to be corrected, and n represents the number of bands.

2. The crosstalk information correction method for a universal mosaic multispectral imaging crop growth sensor device according to claim 1, characterized in that: The S1 includes: S11 fixes the mosaic multispectral imaging crop growth sensor device to the light outlet end of the integrating sphere of the adjustable monochromatic light source system, aligns and parallels the imaging objective lens of the sensor device with the center hole of the integrating sphere, and ensures that the field of view of the imaging objective lens completely covers the inner area of ​​the integrating sphere; S12 connects the computer to the mosaic multi-spectral imaging crop growth sensor device and the adjustable monochromatic light source system, sets the exposure time and gain value of the mosaic multi-spectral imaging crop growth sensor device to 100ms and 6dB, adjusts the adjustable monochromatic light source system according to the band of the mosaic filter, and finds the band with the highest grayscale value in the mosaic filter; under the condition of the band with the highest grayscale value, adjusts the grayscale value of the mosaic multi-spectral imaging crop growth sensor device to 800-1000, and sets the exposure time and gain value at this time as the reference value; S13 uses a computer to adjust the band of the adjustable monochromatic light source system. First, the minimum band value of the adjustable monochromatic light source system is set to be smaller than the minimum band value of the mosaic filter, and the maximum band value is set to be larger than the maximum band value of the mosaic filter. Then, the band of the adjustable monochromatic light source system is adjusted step by step at an interval of 2 nm between the minimum band and the maximum band. S14 uses a computer to control the mosaic multi-spectral imaging crop growth sensor device to collect uniform light source images of each band that are transmitted to the integrating sphere after the step-by-step splitting of the adjustable monochromatic light source system band.

3. The crosstalk information correction method for a universal mosaic multispectral imaging crop growth sensor device according to claim 1, characterized in that: The mosaic multispectral imaging crop growth sensor device includes: An imaging objective lens (1), a mosaic multispectral camera 1 (2), a mosaic multispectral camera 2 (3), a front fixing device (4), a main control system fixing part (5), a main control system (6) and a cover (7), wherein the main control system fixing part (5) fixes the main control system (6) at the rear end of the cover (7); the front fixing device (4) fixes the imaging objective lens (1), the mosaic multispectral camera 1 (2) and the mosaic multispectral camera 2 (3) at the front end of the cover (7) in sequence, and light passes through the imaging objective lens (1) and is received by the mosaic multispectral camera 1 (2) and the mosaic multispectral camera 2 (3) in sequence, and the mosaic multispectral camera 1 (2) and the mosaic multispectral camera 2 (3) are connected to the main control system (6) to send data thereto; and the main control system (6) performs real-time online processing and interpretation of the received spectrum information.

Images