Vegetation area index measurement method and system based on branch dip angle extinction principle

Abstract

The invention relates to the technical field of ecological monitoring, in particular to a vegetation area index measurement method and system based on a branch dip angle extinction principle, and is suitable for ecological monitoring of sparse vegetation in arid and semi-arid regions. The method comprises the following steps: setting a shooting environment and collecting images; cutting the collected image, and performing segmentation through a maximum likelihood method to generate a leaf and xylem binary image; the leaf porosity and the wood porosity are calculated based on the binary image, and leaf parts and branch parts are distinguished by combining spectral characteristics; and respectively calculating a leaf area index and a wood area index through a Markov model and a Beer law, and superposing and inverting a vegetation area index. According to the method, the applicability of the vegetation index in a complex shrub scene is effectively improved. According to the method, high-precision measurement of the vegetation area index in the arid region is realized, the defects of a traditional method in lignified vegetation treatment are overcome, and an important technical support is provided for desertification monitoring.

Description

Technical Field

[0001] The invention relates to the technical field of ecological monitoring, and in particular to a vegetation area index measurement method and system based on the branch inclination angle extinction principle. Background Art

[0002] The vegetation area index (PAI) is an important indicator for evaluating the ecological function of vegetation and the distribution of water resources. In arid and semi-arid areas, sparsely distributed desert shrubs (such as Haloxylon ammodendron) have a significant response to precipitation and soil moisture. PAI can indirectly monitor the degree of desertification and provide a scientific basis for desertification control. However, the special morphology of vegetation in arid areas (such as a high proportion of coniferous and lignified vegetation) makes it difficult for traditional inversion algorithms to accurately describe its ecological characteristics.

[0003] Traditional methods are mostly developed based on the optical properties of broad-leaved forests, and PAI inversion is achieved mainly through the following two approaches: an inversion algorithm based on porosity, which uses the finite length averaging method (Lang and Xiang, 1986) to apply Beer's law on randomly distributed leaves for measurement; an inversion algorithm based on pore size distribution to eliminate large pores in the canopy (Chen and Cihlar, 1995).

[0004] However, these methods have significant defects when dealing with desert shrubs: the influence of xylem on radiation transmission is ignored, resulting in deviations in the calculation results; instruments (such as LAI-2000 and TRAC) have difficulty capturing the details of complex vegetation structures; and the mixed pixel effect in porosity estimation is particularly prominent in sparse shrub environments. Summary of the invention

[0005] In view of the many problems existing in the above-mentioned prior art, the present invention provides a vegetation area index measurement method and system based on the branch inclination extinction principle. The present invention is based on the plant canopy radiation transmission model and the branch inclination extinction principle, combined with high-resolution image data collected by digital photography, and uses the maximum likelihood method to perform color segmentation on the leaf and xylem areas. The leaf area index (LAI) and the wood area index (WAI) are inverted by combining the Markov model and Beer's law, and finally the vegetation area index (PAI) is calculated.

[0006] like Figure 2 As shown, a method for measuring vegetation area index based on the branch inclination extinction principle includes the following steps:

[0007] Setting up the shooting environment and acquiring images containing the samples to be tested;

[0008] The collected image is cropped, and the color segmentation of the cropped image is performed using the maximum likelihood method based on the marked points in the image and the top of the sample, so as to generate binary images of the leaf area and the xylem area;

[0009] Based on binary images, leaf porosity and wood porosity were calculated by pixel statistics, and the leaf area and branch area were distinguished by combining spectral characteristics.

[0010] The leaf area index and wood area index were calculated respectively by the canopy radiation transfer algorithm based on the Markov model and Beer's law, and the vegetation area index was inverted based on the superposition relationship between the leaf area index and the wood area index.

[0011] Preferably, setting the shooting environment includes using a blue background to enhance the separation effect of the wood part and the ground, and fixing the camera on a tripod during the shooting process.

[0012] Preferably, the setting of the marking points during the cropping process uses 1.2 meters as a standard length to calibrate the proportional relationship and spatial scale of the image.

[0013] Preferably, the maximum likelihood method is used to classify the spectral features in the image, and generate binary images of the leaf region and the xylem region based on the color distribution difference.

[0014] Preferably, pixel statistics include calculating the pixel ratio of leaf porosity and wood porosity in the binary image, and further refining the classification accuracy of the leaf area and the branch area in combination with the spectral characteristics.

[0015] Preferably, the canopy radiation transfer algorithm is based on the assumption of the Markov model, regards the canopy as a uniform medium, and uses Beer's law to calculate the radiation attenuation characteristics of the leaf area index and the wood area index.

[0016] Preferably, the vegetation area index is calculated by superimposing the leaf area index and the wood area index, and is defined as half of the sum of the total leaf surface area and the wood surface area per unit ground surface area.

[0017] Preferably, the zenith angles of the images taken with different zenith angles are 55°, 57.5° and 60°, respectively, preferably 57.5°.

[0018] Preferably, the mean is obtained by sampling the axis length ratio of branches of a specific type of plants, and the mean is determined as the empirical value of the axis length ratio of this type of vegetation.

[0019] like Figure 3 As shown, a system for implementing the vegetation area index measurement method based on the branch inclination angle extinction principle comprises:

[0020] An image acquisition module is used to set up the shooting environment and acquire images containing the sample to be tested. The image acquisition module includes a blue background, marking points, a fixed tripod and a camera with adjustable angle;

[0021] An image processing module is used to crop the collected image, taking the marked points in the image and the top of the sample as the reference, and using the maximum likelihood method to perform color segmentation on the cropped image to generate binary images of the leaf area and the xylem area;

[0022] A data calculation module is used to calculate leaf porosity and wood porosity respectively based on binary images through pixel statistics, and distinguish leaf areas from branch areas in combination with spectral features;

[0023] The index calculation module is used to calculate the leaf area index and the wood area index respectively through the canopy radiation transfer algorithm based on the Markov model and Beer's law, and to invert the vegetation area index based on the superposition relationship between the leaf area index and the wood area index.

[0024] Compared with the prior art, the advantages and beneficial effects of the present invention are:

[0025] The present invention uses digital photography combined with blue background technology to effectively separate the wood from the ground background;

[0026] The present invention distinguishes leaf and branch areas by spectral characteristics, and uses the Markov model and Beer's law to calculate the leaf area index and wood area index respectively;

[0027] The present invention preferably uses images at multiple angles (55°, 57.5°, 60°) to verify the reliability of the model, significantly reducing the impact of mixed pixel effects on the results;

[0028] The present invention uses binary image pixel statistics to significantly improve calculation accuracy and expand the applicability of PAI in sparse bush scenes in arid areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of the implementation of the vegetation area index (PAI) inversion algorithm in the example of the present invention;

[0030] Figure 2 It is a schematic diagram of the process of the present invention;

[0031] Figure 3 It is a structural block diagram of the system of the present invention. DETAILED DESCRIPTION

[0032] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is obvious that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0033] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0034] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0035] Vegetation indices, such as the Plant Area Index (PAI), do not distinguish between green and non-green elements, nor do they distinguish between leaves and other elements. Influenced by the research focus, other vegetation indices need to be considered, such as the Leaf Area Index (LAI), which is defined as half of the total leaf area per unit ground area. Vegetation indices should not only consider green elements, but also count the woody parts. The Woody Area Index (WAI) is usually calculated as half of the total woody surface area per unit ground surface area, including branches and stems. There is a simple relationship between the above vegetation indices, PAI = LAI + WAI. Prior studies often use destructive measurements and allometric relationship measurements as direct measurement methods. Direct measurements are often time-consuming and laborious, so they are only used to measure LAI and provide a reference for the calibration or evaluation of indirect methods as the true LAI. Among the indirect measurement methods, optical methods have become the mainstream method for LAI measurement, and algorithm inversion approaches can be divided into two categories. First, the inversion algorithm based on porosity, for example, the finite length averaging method (Lang and Xiang, 1986) applies Beer's law on a small scale where leaves are nearly randomly distributed. Second, the inversion algorithm based on the combination of porosity and pore size, for example, the pore size distribution method (Chen and Cihlar 1995a) effectively eliminates the large pores between crowns. Subsequent studies are mostly based on further improvements and deformations of the above two approaches. The optical instruments developed therewith, such as DEMON, LAI-2000, TRAC, etc., use small portable sensors as the main sampling equipment, and linear measuring instruments are difficult to discover the fine details of complex vegetation. The algorithm of the present invention reasonably explores the high spatial resolution of digital photography, and benefits from the precise distinction of vegetation structure to improve the accuracy of vegetation area index estimation.

[0036] In the specific implementation process, the present invention uses photogrammetry as the main means. When shooting the sample, considering the camera shooting ratio, a mark point with a length of 1.2m is placed with the sample as the center and it is ensured that the shot image contains the mark point. At the same time, in order to solve the problem that the wood part is difficult to separate from the ground, the indoor shooting background is set to blue, the shooting camera is fixed with a tripod, and the goniometer is placed parallel to the camera lens for reading. The north of the initial position of the sample is placed as the visual direction angle of 0°, the visual direction angle is kept unchanged, and then the sample image is shot while keeping the visual zenith angles of 55°, 57.5°, and 60° unchanged respectively. After shooting, measure the height of the Haloxylon ammodendron tree and record it.

[0037] The captured images were cropped with the marked points and the top of the Haloxylon ammodendron tree as the vertices and edges, and the cropped images were segmented using the maximum likelihood method to obtain the leaf and wood binary images based on the differences in spectral characteristics of different vegetation structures. The leaf porosity and wood porosity were then obtained through pixel statistics. Finally, the inversion algorithm was used to obtain the actual measured prediction value.

[0038] The biological characteristics of Haloxylon ammodendron indicate that the leaves have been severely degraded, with assimilating branches as photosynthetic organs. For ease of understanding, the assimilating branch area index is referred to as the leaf area index (LAI) and the assimilating branches are referred to as leaves. In this study, a simple method was used to distinguish branches from leaves by spectral characteristics from summer measurements in the desert. According to the basic principle:

[0039] PAI=LAI+WAI

[0040] Finally, the vegetation area index is obtained. The inversion algorithm process of the vegetation area index is as follows: Figure 1 shown.

[0041] The extinction model based on vegetation structure describes the possibility of radiation interception by different vegetation structures in the canopy. In practical applications, the most commonly used method to estimate leaf interception is to treat the canopy as a uniform medium (i.e., a "turbid" medium) and use a simple radiation attenuation exponential model for estimation, namely, Beer's law. Beer's law calculates the probability of radiation interception as an exponential growth function of the projected leaf area in the direction of radiation propagation and the distance through the canopy, so it is also called the penetration function, which was proposed in the Markov model (Nilson, 1971). It is:

[0042]

[0043] Among them, P l G(θ represents the porosity, which can be understood as the possibility of light spots appearing at the bottom of the canopy when only leaves are considered. v ,θ l ) represents the unit leaf area perpendicular to θ vThe projection coefficient function on the plane of the direction is referred to as the G function. In the case of no preferred azimuthal orientation (isotropic), G can be simplified as follows:

[0044]

[0045] Among them, A(θ v ,θ l ) represents the leaf area projection kernel function, which is given by the following formula:

[0046]

[0047] Where φ = cos -1 (cotθ v cotθ l ).

[0048] G(θ l ) represents the leaf tilt angle distribution function (LAD), which indicates the leaf normal at θ l The probability of directional distribution, that is, the distribution probability density of leaf inclination. In prior studies, due to the neglect of xylem light interception, most distribution functions are studied for leaf distribution, such as de Wit's typical six-type distribution, Beta distribution, Verhoef algorithm, etc. Among them, the ellipsoid distribution function shows better adaptability to describe desert vegetation from both theoretical and actual measurements. It is defined as:

[0049]

[0050] Among them, χ = b / a, which represents the flattening of the ellipsoid.

[0051] Starting from the perspective of canopy structure, the present invention raises mathematical modeling from the usual needles to a higher level structure, such as the spatial distribution pattern of branches and twigs. Since the interception of light by branches is different from that of leaves, a binomial is used to calculate the interception of light by branches. The expression is:

[0052]

[0053] Among them, P s represents the porosity when only the branches are considered. v ,θ s) The unit branch area is perpendicular to θ v The projection coefficient function on the plane of the direction, hereinafter referred to as F function. When isotropy is assumed, it can be simplified as follows (Wang et al., 2007; Wilson, 1960):

[0054]

[0055] Among them, A w (θv ,θ s ) represents the branch area projection kernel function. Since prior studies mostly focus on broad-leaved and deciduous trees, the xylem has little effect, so the xylem projection is not studied separately. However, the influence of xylem on vegetation in desert areas cannot be ignored, so the branch area projection kernel function is proposed, which is given by the following formula:

[0056]

[0057] Among them, x=h / 2r, which represents the axial length ratio of the branch.

[0058] F(θ s ) represents the branch tilt angle distribution function. The typical research object of the present invention, Haloxylon ammodendron, has the biological characteristics of its assimilated branches. It is believed that the branch tilt angle distribution is consistent with the leaf tilt angle distribution. g(θ l ) represents the leaf tilt angle distribution function, that is, f(θ s )=g(θ l ). It is defined as:

[0059]

[0060] Among them, χ=b / a, which represents the flattening of the ellipsoid function.

[0061] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware.

[0062] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for measuring vegetation area index based on the branch inclination extinction principle, characterized in that: The following steps are involved: Setting up the shooting environment and acquiring images containing the samples to be tested; The collected image is cropped, and the color segmentation of the cropped image is performed using the maximum likelihood method based on the marked points in the image and the top of the sample, so as to generate binary images of the leaf area and the xylem area; Based on binary images, leaf porosity and wood porosity were calculated by pixel statistics, and the leaf area and branch area were distinguished by combining spectral characteristics. The leaf area index and wood area index were calculated respectively by the canopy radiation transfer algorithm based on the Markov model and Beer's law, and the vegetation area index was inverted based on the superposition relationship between the leaf area index and the wood area index.

2. The method for measuring vegetation area index based on the branch inclination extinction principle according to claim 1, characterized in that: The shooting environment was set up by using a blue background to enhance the separation between the wood and the ground, and the camera was fixed on a tripod during the shooting.

3. The method for measuring vegetation area index based on the branch inclination extinction principle according to claim 1, characterized in that: During the cropping process, the marker points are set with a standard length of 1.2 meters to calibrate the proportional relationship and spatial scale of the image.

4. The method for measuring vegetation area index based on the branch inclination angle extinction principle according to claim 1, characterized in that: The maximum likelihood method is used to classify the spectral features in the image, and binary images of the leaf area and the xylem area are generated based on the differences in color distribution.

5. The method for measuring vegetation area index based on the branch inclination extinction principle according to claim 1, characterized in that: Pixel statistics include calculating the pixel ratio of leaf porosity and wood porosity in binary images, and combining spectral features to further refine the classification accuracy of leaf and branch areas.

6. The method for measuring vegetation area index based on the branch inclination angle extinction principle according to claim 1, characterized in that: The canopy radiation transfer algorithm is based on the assumption of the Markov model, which regards the canopy as a uniform medium and uses Beer's law to calculate the radiation attenuation characteristics of the leaf area index and the wood area index.

7. The method for measuring vegetation area index based on the branch inclination angle extinction principle according to claim 1, characterized in that: The vegetation area index was calculated by superimposing the leaf area index and the wood area index, and was defined as half of the sum of the total leaf surface area and the wood surface area per unit ground surface area.

8. The method for measuring vegetation area index based on the branch inclination extinction principle according to claim 1, characterized in that: The zenith angles of the images taken with different zenith angles are 55°, 57.5° and 60° respectively, with 57.5° being preferred.

9. The method for measuring vegetation area index based on the branch inclination extinction principle according to claim 1, characterized in that: The mean value is obtained by sampling the branch axis length ratio of a specific type of plants, and the mean value is determined as the empirical value of the axis length ratio of this type of vegetation.

10. A system for implementing the vegetation area index measurement method based on the branch inclination angle extinction principle as claimed in any one of claims 1 to 9, characterized in that: The system includes: An image acquisition module is used to set up the shooting environment and acquire images containing the sample to be tested. The image acquisition module includes a blue background, marking points, a fixed tripod and a camera with adjustable angle; An image processing module is used to crop the collected image, taking the marked points in the image and the top of the sample as the reference, and using the maximum likelihood method to perform color segmentation on the cropped image to generate binary images of the leaf area and the xylem area; A data calculation module is used to calculate leaf porosity and wood porosity respectively based on binary images through pixel statistics, and distinguish leaf areas from branch areas in combination with spectral features; The index calculation module is used to calculate the leaf area index and the wood area index respectively through the canopy radiation transfer algorithm based on the Markov model and Beer's law, and to invert the vegetation area index based on the superposition relationship between the leaf area index and the wood area index.

Images