A SIF splitting method based on multispectral camera and spectrometer

Abstract

The present invention discloses a SIF splitting method based on a multispectral camera and a spectrometer, comprising the following steps: calculating the normalized vegetation index NDVI of each pixel according to the pixel brightness value of each pixel of an RGB picture and an NIR picture; calculating the SIF value of the observation area according to vegetation spectrum data and solar spectrum data; matching the spectrometer observation area with the shooting area of ​​the multispectral camera; selecting a pixel area with an NDVI of 0.2-1.0 in the observation area, counting it as a vegetation area, calculating the NDVI of the average vegetation area pixel point, and obtaining the vegetation average NDVI of the observation area; according to the vegetation average NDVI and SIF value and the NDVI of each pixel, the SIF of each pixel point is estimated by a formula: mapping the SIF of each pixel in the vegetation area to the picture, eliminating the non-vegetation area, and obtaining a SIF distribution image. The present invention can distinguish the average SIF measured by the spectrometer at a relatively low cost, and can calculate and compare the SIFs of different areas and different vegetation types in the observation area.

Description

Technical Field

[0001] The invention relates to the field of ecological remote sensing, and in particular to a SIF splitting method based on a multispectral camera and a spectrometer. Background Art

[0002] Solar / Sun-Induced Chlorophyll Fluorescence (SIF) is a type of fluorescence produced by plants under natural light. It is highly correlated with photosynthesis and is considered an effective tool for detecting photosynthesis. Currently, researchers widely use SIF to track photosynthesis dynamics and monitor vegetation stress, which is of great significance for ecosystem carbon sink assessment and environmental monitoring.

[0003] Since the discovery of chlorophyll fluorescence, the research and detection of vegetation chlorophyll fluorescence in the ecological field has never stopped. With the emergence and popularization of hyperspectral technology, breakthrough progress has been made in the research of SIF detection at the canopy scale. Hyperspectral technology can capture the fluorescence spectrum emitted by plants under natural light, providing more real and dynamic photosynthesis information. At present, hyperspectral technology has been widely used in SIF monitoring in the ecological field. By tracking the dynamic changes of photosynthesis in real time at the canopy scale, it makes up for the shortcomings of laser-induced fluorescence detection. In general, the application of hyperspectral technology in SIF detection has not only promoted the development of ecological research, but also provided important tools for environmental monitoring, climate change research and agricultural management. This technological advancement enables us to more accurately assess the health status and photosynthesis efficiency of vegetation, thereby providing a scientific basis for the protection and management of ecosystems.

[0004] SIF measurement equipment is mainly divided into two types: fiber optic and imaging. Fiber Optic Fluorometer uses fiber optic probes to collect the fluorescence spectrum emitted by plants, which can provide high-precision measurement data and is very suitable for laboratory research and small-scale field measurements. At the same time, the cost is relatively low compared to imaging spectrometers, and automatic continuous observation can be achieved. At present, SIF applications on the market are more composed of fiber optic spectrometers. In addition to fiber optic spectrometers, imaging spectrometers have a certain market. They can simultaneously obtain fluorescence images of large areas, and can spatially distinguish SIF values ​​at different locations, thereby realizing the monitoring of large areas of vegetation. It is very suitable for large-scale ecological research and agricultural monitoring.

[0005] Although fiber optic spectrometers can provide high-precision measurement data, their main limitation is that they can only provide average data within the fiber optic field of view, which means that it is difficult for the spectrometer to distinguish the SIF data of different vegetation in the observation area; secondly, if there are other shadows of obstructions in the observation area, it will also affect the SIF value. SIF is a fluorescence induced by sunlight. Reduced light will cause low SIF, but the spectrometer cannot distinguish between light and shadow positions. Although imaging spectrometers can obtain fluorescence images of large areas and spatially distinguish SIF values ​​at different locations, their complex optical systems and data processing requirements increase the purchase and maintenance costs of the equipment, which is very limited in practical applications. Summary of the invention

[0006] The object of the present invention is to provide a SIF splitting method based on a multispectral camera and a spectrometer, so as to solve the above-mentioned problems existing in the prior art.

[0007] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] A SIF splitting method based on a multispectral camera and a spectrometer comprises the following steps:

[0009] S1, calculate the normalized vegetation index NDVI of each pixel point according to the pixel brightness value of each pixel in the RGB image and the NIR image;

[0010] S2, calculate the SIF value of the observation area based on the vegetation spectrum data and the solar spectrum data;

[0011] S3, matching the observation area of ​​the spectrometer with the shooting area of ​​the multi-spectral camera;

[0012] S4, selecting a pixel area with an NDVI of 0.2-1.0 in the observation area, counting it as a vegetation area, calculating the average NDVI of the pixels in the vegetation area, and obtaining the average NDVI of the vegetation in the observation area;

[0013] S5. According to the average NDVI and SIF values ​​of vegetation and the NDVI of each pixel, the SIF of each pixel is estimated by the formula:

[0014] S6. Map the SIF of each pixel in the vegetation area to the image, remove the non-vegetation area, and obtain a SIF distribution image.

[0015] In some specific embodiments, the NDVI calculation formula is as follows:

[0016]

[0017] Where: NDVI pixelNDVI, NIR calculated for each pixel DN is the DN value of each pixel in the NIR band, R DN It is the DN value of each pixel in the R band of the RGB image;

[0018] After the NDVI of each pixel is calculated, it can be mapped onto the original image to form an NDVI image.

[0019] In some specific embodiments, the SIF formula is calculated as follows:

[0020]

[0021] Where: SIF mean is the average SIF in the observation area; L in is the vegetation spectrum data of the dark line band; E out is the solar spectrum data outside the dark line band; L out E is the vegetation spectral data of the band outside the dark line; in It is the solar spectrum data of the dark line band.

[0022] In some specific embodiments, the specific method of step S3 is: the multispectral camera and the spectrometer are installed at the same point and point to the same angle. According to the field of view of the spectrometer and the multispectral camera, a circular spectral observation area with the center of the image as the origin can be calculated on the image area captured by the multispectral camera. The formula is as follows:

[0023]

[0024] Where: P x The pixel radius of the circular spectral observation area with the image center as the origin; P H is the number of horizontal pixels of the image; θ x is the field of view of the spectrometer; θ H is the horizontal field of view of the camera;

[0025] Crop the NDVI image according to the circular area to retain the spectrometer observation area.

[0026] In some specific embodiments, the calculation formula of the SIF of each pixel in step S5 is:

[0027]

[0028] Where: SIFpixel represents the SIF of a single pixel; SIFmean represents the SIF value of the observed area; NDVIpixel represents the NDVI of the pixel in the vegetation area; and NDVImean represents the average NDVI of the vegetation in the observed area.

[0029] In some specific embodiments, in step S6, the color of each pixel represents the SIF value of the pixel, and the average SIF in the selected area can be obtained by selecting an area on the image and averaging the SIF values ​​of the pixels in the area.

[0030] In some specific embodiments, it further includes: a splitting device used in the splitting method;

[0031] The split equipment includes: multispectral camera and spectrometer;

[0032] The spectrometer includes two spectrum collection optical fibers, which are respectively pointed to the vegetation and the sky. The optical fiber pointed to the vegetation can determine the position of the spectrometer observation area according to the optical fiber field of view and the observation height;

[0033] The multispectral camera and spectrometer are installed at the same point and point to the same angle to capture vegetation images. The field of view of the multispectral camera is larger than that of the optical fiber, so that the multispectral camera covers the observation area of ​​the spectrometer. During observation, the multispectral camera and spectrometer acquire data synchronously. The multispectral camera captures color (RGB) and near-infrared (NIR) band images of the observation area, and the spectrometer acquires vegetation spectral data and solar spectral data of the observation area.

[0034] The beneficial effects of the present invention are as follows: the present invention discloses a SIF splitting method based on a multispectral camera and a spectrometer, comprising the following steps: calculating the normalized vegetation index NDVI of each pixel point according to the pixel brightness value of each pixel in the RGB image and the NIR image; calculating the SIF value of the observation area according to the vegetation spectrum data and the solar spectrum data; matching the spectrometer observation area with the shooting area of ​​the multispectral camera; selecting a pixel area with an NDVI of 0.2-1.0 in the observation area, counting it as the vegetation area, calculating the NDVI of the average vegetation area pixel point, and obtaining the vegetation average NDVI of the observation area; estimating the SIF of each pixel point according to the vegetation average NDVI and the SIF value and the NDVI of each pixel by the formula: mapping the SIF of each pixel in the vegetation area to the image, eliminating the non-vegetation area, and obtaining the SIF distribution image. The SIF splitting method based on a multispectral camera proposed in this scheme of the present invention can split the SIF observed by the spectrometer according to the spectrum image observed in real time by the multispectral camera, and obtain the fluorescence distribution image of the SIF in the observation area. Through this method, the average SIF measured by the spectrometer can be distinguished at a lower cost. In practical applications, the SIFs of different areas and different vegetation types in the observation area can be calculated and compared. By selecting the illuminated area, the influence of occlusion shadows on the vegetation SIF measurement can also be eliminated to achieve shadow correction of the vegetation SIF. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of a SIF splitting method based on a multispectral camera and a spectrometer of the present invention;

[0036] Figure 2 It is a data observation schematic diagram of the present invention;

[0037] Figure 3 is a pixel-level SIF estimation flow chart of the present invention;

[0038] Figure 4 is an RGB image taken by the multispectral camera of the present invention;

[0039] Figure 5 is a NIR image taken by the multispectral camera of the present invention;

[0040] Figure 6 It is the NDVI image calculated from the RGB and NIR images of the present invention;

[0041] Figure 7 It is the NDVI image of the spectrum observation area of ​​the present invention;

[0042] Figure 8 It is a schematic diagram of identifying non-vegetation areas of the present invention;

[0043] Fig. 9 It is a schematic diagram of the SIF calculated by the present invention after being mapped to an image. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation methods described herein are only used to explain the present invention and are not used to limit the present invention.

[0045] Reference Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 7 , Figure 8 and Fig. 9 A SIF splitting method based on a multispectral camera and a spectrometer is shown, comprising the following steps:

[0046] S1, calculate the normalized vegetation index NDVI of each pixel point according to the pixel brightness value of each pixel in the RGB image and the NIR image;

[0047] S2, calculate the SIF value of the observation area based on the vegetation spectrum data and the solar spectrum data;

[0048] S3, matching the observation area of ​​the spectrometer with the shooting area of ​​the multi-spectral camera;

[0049] S4. Select pixel areas with NDVI of 0.2-1.0 in the observation area and count them as vegetation areas. Calculate the NDVI of the average vegetation area pixels to obtain the average vegetation NDVI of the observation area. The typical range of vegetation NDVI is usually between 0.2 and 1.0, reflecting different degrees of vegetation coverage and health: values ​​below 0.2 indicate areas without vegetation such as bare soil, water bodies or buildings.

[0050] S5. Estimate the SIF of each pixel using the formula based on the vegetation average NDVI and SIF values ​​and the NDVI of each pixel. It should be noted that based on the correlation between vegetation NDVI and SIF, the NDVI of each pixel is proportional to its SIF contribution.

[0051] S6. Map the SIF of each pixel in the vegetation area to the image, remove the non-vegetation area, and obtain a SIF distribution image.

[0052] In this embodiment, the Normalized Difference Vegetation Index (NDVI) of vegetation and SIF have a high correlation, so the NDVI of each corresponding pixel is calculated based on the image taken by the multispectral camera, and the SIF is mapped to the pixel point of the multispectral image.

[0053] In some specific embodiments, the NDVI calculation formula is as follows:

[0054]

[0055] Where: NDVI pixel NDVI, NIR calculated for each pixel DN is the DN value of each pixel in the NIR band, R DN It is the DN value of each pixel in the R band of the RGB image;

[0056] After the NDVI of each pixel is calculated, it can be mapped onto the original image to form an NDVI image.

[0057] In some specific embodiments, the SIF formula is calculated as follows:

[0058]

[0059] Where: SIF mean is the average SIF in the observation area; L in is the vegetation spectrum data of the dark line band; E out is the solar spectrum data outside the dark line band; L out E is the vegetation spectral data of the band outside the dark line; in It is the solar spectrum data of the dark line band.

[0060] In some specific embodiments, the specific method of step S3 is: the multispectral camera and the spectrometer are installed at the same point and point to the same angle. According to the field of view of the spectrometer and the multispectral camera, a circular spectral observation area with the center of the image as the origin can be calculated on the image area captured by the multispectral camera. The formula is as follows:

[0061]

[0062] Where: P x The pixel radius of the circular spectral observation area with the image center as the origin; P H is the number of horizontal pixels of the image; θ x is the field of view of the spectrometer; θ H is the horizontal field of view of the camera;

[0063] Crop the NDVI image according to the circular area to retain the spectrometer observation area.

[0064] In some specific embodiments, the calculation formula of the SIF of each pixel in step S5 is:

[0065]

[0066] Where: SIFpixel represents the SIF of a single pixel; SIFmean represents the SIF value of the observed area; NDVIpixel represents the NDVI of the pixel in the vegetation area; and NDVImean represents the average NDVI of the vegetation in the observed area.

[0067] In some specific embodiments, in step S6, the color of each pixel represents the SIF value of the pixel, and the average SIF in the selected area can be obtained by selecting an area on the image and averaging the SIF values ​​of the pixels in the area.

[0068] Through the above method, the SIF value of the observation area can be split to obtain the SIF values ​​of different areas and different vegetation (grass or trees, etc.); at the same time, the influence of shadow occlusion on vegetation SIF observation can be corrected. Due to the reduction of light in the shadow, the near-infrared light reflection is greater than the red light, and the NDVI is also reduced. This is consistent with the physical law of SIF under light and shadow. Therefore, the SIF value of each pixel calculated by the NDVI weight can reflect the SIF distribution law under light and shadow. By selecting and calculating the SIF mean of the illuminated area instead of the regional SIF observed by the spectrometer, the shadow correction of SIF can be achieved.

[0069] In some specific embodiments, it further includes: a splitting device used in the splitting method;

[0070] The split equipment includes: multispectral camera and spectrometer;

[0071] The spectrometer includes two spectrum collection optical fibers, which are respectively pointed to the vegetation and the sky. The optical fiber pointed to the vegetation can determine the position of the spectrometer observation area according to the optical fiber field of view and the observation height;

[0072] The multispectral camera and spectrometer are installed at the same point and point to the same angle to capture vegetation images. The field of view of the multispectral camera is larger than that of the optical fiber, so that the multispectral camera covers the observation area of ​​the spectrometer. During observation, the multispectral camera and spectrometer acquire data synchronously. The multispectral camera captures color (RGB) and near-infrared (NIR) band images of the observation area, and the spectrometer acquires vegetation spectral data and solar spectral data of the observation area.

[0073] The beneficial effects of the present invention are as follows: the present invention discloses a SIF splitting method based on a multispectral camera and a spectrometer, comprising the following steps: calculating the normalized vegetation index NDVI of each pixel point according to the pixel brightness value of each pixel in the RGB image and the NIR image; calculating the SIF value of the observation area according to the vegetation spectrum data and the solar spectrum data; matching the spectrometer observation area with the shooting area of ​​the multispectral camera; selecting a pixel area with an NDVI of 0.2-1.0 in the observation area, counting it as the vegetation area, calculating the NDVI of the average vegetation area pixel point, and obtaining the vegetation average NDVI of the observation area; estimating the SIF of each pixel point according to the vegetation average NDVI and the SIF value and the NDVI of each pixel by the formula: mapping the SIF of each pixel in the vegetation area to the image, eliminating the non-vegetation area, and obtaining the SIF distribution image. The SIF splitting method based on a multispectral camera proposed in this scheme of the present invention can split the SIF observed by the spectrometer according to the spectrum image observed in real time by the multispectral camera, and obtain the fluorescence distribution image of the SIF in the observation area. Through this method, the average SIF measured by the spectrometer can be distinguished at a lower cost. In practical applications, the SIFs of different areas and different vegetation types in the observation area can be calculated and compared. By selecting the illuminated area, the influence of occlusion shadows on the vegetation SIF measurement can also be eliminated to achieve shadow correction of the vegetation SIF.

Claims

1. A SIF splitting method based on a multispectral camera and a spectrometer, characterized in that: The following steps are involved: S1, calculate the normalized vegetation index NDVI of each pixel point according to the pixel brightness value of each pixel in the RGB image and the NIR image; S2, calculate the SIF value of the observation area based on the vegetation spectrum data and the solar spectrum data; S3, matching the observation area of ​​the spectrometer with the shooting area of ​​the multi-spectral camera; S4, selecting a pixel area with an NDVI of 0.2-1.0 in the observation area, counting it as a vegetation area, calculating the average NDVI of the pixels in the vegetation area, and obtaining the average NDVI of the vegetation in the observation area; S5. Estimate the SIF of each pixel point according to the average NDVI and SIF values ​​of the vegetation and the NDVI of each pixel using the formula: S6, mapping the SIF of each pixel in the vegetation area to the image, removing the non-vegetation area, and obtaining a SIF distribution image; The calculation formula of the SIF of each pixel in step S5 is: Where: SIF pixel SIF represents a single pixel; SIF mean Indicates the SIF value of the observation area; NDVI pixel Indicates the NDVI of the pixel point in the vegetation area; NDVI mean Represents the average NDVI of vegetation in the observation area.

2. The SIF splitting method based on a multispectral camera and a spectrometer according to claim 1, characterized in that: The NDVI calculation formula is as follows: Where: NDVI pixel NDVI, NIR calculated for each pixel DN is the DN value of each pixel in the NIR band, R DN It is the DN value of each pixel in the R band of the RGB image; After the NDVI of each pixel is calculated, it is mapped onto the original image to form an NDVI image.

3. The SIF splitting method based on a multispectral camera and a spectrometer according to claim 1, characterized in that: The SIF formula in step 2 is calculated as follows: Where: SIF mean is the average SIF in the observation area; L in is the vegetation spectrum data in the dark line band; E out is the solar spectrum data outside the dark line band; L out E is the vegetation spectral data of the band outside the dark line; in It is the solar spectrum data of the dark line band.

4. The SIF splitting method based on a multispectral camera and a spectrometer according to claim 1, characterized in that: The specific method of step S3 is: the multispectral camera and the spectrometer are installed at the same point and pointed at the same angle. According to the field of view of the spectrometer and the multispectral camera, a circular spectral observation area with the center of the image as the origin can be calculated on the image area captured by the multispectral camera. The formula is as follows: Where: P x The pixel radius of the circular spectral observation area with the image center as the origin; P H is the number of horizontal pixels of the image; θ x is the field of view of the spectrometer; θ H is the horizontal field of view of the camera; Crop the NDVI image according to the circular area to retain the spectrometer observation area.

5. The SIF splitting method based on a multispectral camera and a spectrometer according to claim 1, characterized in that: In step S6, the color of each pixel represents the SIF value of the pixel. By selecting a region on the image and averaging the SIF values ​​of the pixels in the region, the average SIF in the selected region can be obtained.

6. The SIF splitting method based on a multispectral camera and a spectrometer according to claim 1, characterized in that: Also included: a splitting device used for the splitting method; The splitting equipment includes: a multi-spectral camera and a spectrometer; The spectrometer includes two spectrum collection optical fibers, which are respectively pointed to the vegetation and the sky. The optical fiber pointed to the vegetation can determine the position of the spectrometer observation area according to the optical fiber field of view and the observation height; The multispectral camera and the spectrometer are installed at the same point, pointing to the same angle, to capture vegetation images. The field of view of the multispectral camera is larger than that of the optical fiber, so that the multispectral camera covers the observation area of ​​the spectrometer. During observation, the multispectral camera and the spectrometer synchronously acquire data. The multispectral camera captures color RGB band and near-infrared NIR band images of the observation area, and the spectrometer acquires vegetation spectrum data and solar spectrum data of the observation area.

Images