A method for target detection and recognition in complex background

Abstract

The present invention relates to the technical field of optical detection and target identification, and specifically discloses a method for detecting and identifying targets in a complex background, comprising: determining an application scenario to obtain scene simulation parameters; constructing a three-dimensional simulated scene using 3D modeling software; simulating the spectral characteristics, scattering intensity, and polarization characteristics of a specific target and multiple complex backgrounds using a background model; performing scene simulation based on the scene simulation parameters, the three-dimensional simulated scene, and the background model, and obtaining a simulated grayscale image; performing grayscale comparison between the simulated grayscale image and a measured image of the application scenario, and detecting and identifying a specific target whose presence or absence in the application scenario is uncertain using a target recognition algorithm, thereby obtaining a simulated scene and target recognition results. The present invention uses a spectral detection method to improve the recognition probability of a target in a complex background.

Description

Technical Field

[0001] The present invention relates to the technical field of optical detection and recognition of targets, and in particular to a method for detecting and recognizing targets under complex backgrounds. Background Art

[0002] With the development of science and technology, target recognition technology has been widely used in fields such as biomedicine, satellite remote sensing, robotic vision, cargo inspection, target tracking, autonomous vehicle navigation, banking, transportation, military, e-commerce, and multimedia network communications. Existing target recognition methods, mainly deep learning methods, are the mainstream recognition method. These methods usually rely on a large amount of data samples to support recognition accuracy. Existing target recognition methods, especially when identifying targets of the same color in complex backgrounds, generally suffer from poor anti-interference and low recognition rate.

[0003] Therefore, those skilled in the art urgently need to provide a new target recognition method under complex backgrounds to overcome the defects in the above-mentioned prior art. Summary of the Invention

[0004] Therefore, the technical problem to be solved by the present invention is to overcome the defects existing in the prior art, thereby providing a method for target detection and identification under complex backgrounds by combining Bidirectional Reflectance Distribution Function, a bidirectional reflectance distribution function, referred to as BRDF function, which is an important function used to describe the reflection characteristics of the object surface to the incident light, to perform target detection and identification under complex backgrounds.

[0005] A method for detecting and identifying targets in a complex background comprises the following steps:

[0006] Determine the application scenario and obtain scenario simulation parameters;

[0007] Use 3D modeling software to process scene simulation parameters and construct a three-dimensional simulation scene;

[0008] Use pre-built background models to simulate the spectral characteristics, scattering intensity, and polarization characteristics of specific targets and various complex backgrounds;

[0009] Performing scene simulation according to scene simulation parameters, three-dimensional simulation scene, and background model, and further acquiring a simulated grayscale image;

[0010] The grayscale of the simulated grayscale image and the measured image of the application scene are compared, and the specific target that is uncertain whether it exists in the application scene is identified and detected through the target recognition algorithm to obtain the simulation scene and target recognition results.

[0011] Preferably, the simulation process also includes simulation modeling of the application scene based on the reflectivity of the specific target material itself and the spatial coordinates of the three-dimensional simulation scene.

[0012] Preferably, obtaining a simulated grayscale image specifically includes the following steps:

[0013] The sun's incident direction is introduced for spatial calculation, and the full-space BRDF value of a specific target is obtained by traversing the entire space according to the BRDF function;

[0014] Set the observation direction, calculate the BRDF value of each facet of the specific target material at the corresponding angle, and obtain a simulated grayscale image of the simulated scene.

[0015] Preferably, the multiple complex backgrounds specifically include: woodland, desert, grassland, road and snow.

[0016] Preferably, the background model includes the following corresponding to various complex backgrounds: a forest pBRDF model, a sand pBRDF model, a grassland pBRDF model, a highway pBRDF model, and a snow BRDF model;

[0017] Among them, the snow BRDF model expression is:

[0018] ;

[0019] formula, Represents the specular reflection component of the sample surface BRDF; is the normal distribution function of the small facets on the sample surface, is the Fresnel approximation function, is the masking function; Indicates the BRDF value of the snow model; 、 、 , b are parameters to be determined;

[0020] Specifically: and Reflects the magnitude of the specular reflection and diffuse reflection components respectively, and is related to the roughness and reflectivity of the target surface; b is the parameter reflecting the Fresnel reflection coefficient of the target surface and is related to the refractive index of the target. are the incident zenith angle, incident azimuth angle, scattering zenith angle and scattering azimuth angle respectively; It represents the angle between the ray from the point on the aperture to the observation point and the normal of the aperture; Represents the angle between the target surface normal and the microfacet normal.

[0021] Preferably, the sand pBRDF model expression is:

[0022] ;

[0023] ;

[0024] in, is the sand mirror reflection component, is the normal distribution function, θ represents the angle between the incident light and the reflected light; σ represents the surface roughness; is the masking factor, Indicates relative azimuth; is the degree of polarization, represents the refractive index, is a 4×4 Fresnel reflection-Mueller matrix; A is the fitting parameter, β represents the angle between the reflection direction and the normal direction, B is the fitting parameter, represents the reflected S wave, represents the reflected P wave, Indicates the BRDF value of the sand model.

[0025] Preferably, the grass pBRDF model expression is:

[0026] ;

[0027] in, represents the slope variance of the rough surface; is the specular reflection coefficient; is the reflectivity of the target surface; is the free scattering coefficient; represents the free fitting parameter, represents a six-parameter model; represents the depolarization matrix for volume scattering, represents the depolarization matrix of backscattering; represents the fitting factor.

[0028] Preferably, the forestland pBRDF model expression is:

[0029] = ；

[0030] ;

[0031] ;

[0032] Where: represents the forestland pBRDF model, represents the diffuse component of the forest floor pBRDF model, is the surface autocorrelation length function, Represents the BRDF value of the specular reflection component.

[0033] Preferably, the highway pBRDF model expression is:

[0034] .

[0035] The technical solution of the present invention has the following advantages:

[0036] The method of spectral detection is used to improve the recognition probability of targets in complex backgrounds, and at the same time, the BRDF scattering of light of targets in all directions in complex backgrounds is revealed, and a research idea is provided for target detection and recognition in complex backgrounds. In addition, the method of the present invention simulates the spectral characteristics and scattering intensity of targets in five typical complex backgrounds: woodland, desert, grassland, road and snow, and provides technical inspiration for other complex backgrounds with practical needs. It improves the recognition probability of targets in complex backgrounds, and at the same time, the bidirectional reflection distribution of light of targets in all directions in complex backgrounds is revealed, and a research idea is provided for target detection and recognition in complex backgrounds. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 The grassland background modeling diagram of the present invention;

[0039] Figure 2 A BRDF polar coordinate diagram corresponding to the grass background modeling diagram of the present invention;

[0040] Figure 3 A forest background modeling diagram of the present invention;

[0041] Figure 4 A BRDF polar coordinate diagram corresponding to the woodland background modeling diagram of the present invention;

[0042] Figure 5 The sandy land background modeling diagram of the present invention;

[0043] Figure 6 The BRDF polar coordinate diagram corresponding to the sand background modeling diagram of the present invention;

[0044] Figure 7 This is the snow background modeling diagram of the present invention;

[0045] Figure 8 A BRDF polar coordinate diagram corresponding to the snow background modeling diagram of the present invention;

[0046] Figure 9 The highway background modeling diagram of the present invention;

[0047] Figure 10 A BRDF polar coordinate diagram corresponding to the highway background modeling diagram of the present invention;

[0048] Figure 11 This is a schematic diagram of the jeep modeling results of the present invention;

[0049] Figure 12 A schematic diagram showing the modeling details of the jeep of the present invention;

[0050] Figure 13 This is a target simulation diagram under the forest background of the present invention;

[0051] Figure 14 This is a schematic diagram of target simulation results before camouflage according to the present invention;

[0052] Figure 15 This is a schematic diagram of the simulation result of the camouflaged target of the present invention;

[0053] Figure 16 The target of the present invention and the spectral reflectance curve of grassland are shown;

[0054] Figure 17 This is a schematic diagram of target simulation results under a grassland background according to the present invention;

[0055] Figure 18 This is a flow chart of a target detection and recognition method under complex backgrounds according to the present invention. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0057] In the description of the present invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0058] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0059] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0060] Example 1:

[0061] like Figure 18 This embodiment discloses a method for detecting and identifying targets in a complex background, comprising the following steps:

[0062] Determine the application scenario and obtain scenario simulation parameters;

[0063] Use 3D modeling software to process scene simulation parameters and construct a three-dimensional simulation scene;

[0064] Use pre-built background models to simulate the spectral characteristics, scattering intensity, and polarization characteristics of specific targets and various complex backgrounds;

[0065] Performing scene simulation according to scene simulation parameters, three-dimensional simulation scene, and background model, and further acquiring a simulated grayscale image;

[0066] The grayscale of the simulated grayscale image and the measured image of the application scene are compared, and the specific target that is uncertain whether it exists in the application scene is identified and detected through the target recognition algorithm to obtain the simulation scene and target recognition results.

[0067] Specifically:

[0068] The simulation process also includes simulation modeling of the application scene based on the reflectivity of the specific target material itself and the spatial coordinates of the three-dimensional simulation scene.

[0069] In this embodiment, obtaining a simulated grayscale image specifically includes the following steps:

[0070] The sun's incident direction is introduced for spatial calculation, and the full-space BRDF value of a specific target is obtained by traversing the entire space according to the BRDF function;

[0071] Set the observation direction, calculate the BRDF value of each facet of the specific target material at the corresponding angle, and obtain a simulated grayscale image of the simulated scene.

[0072] The pre-built background models include the following corresponding to various complex backgrounds: forest pBRDF model, sand pBRDF model, grassland pBRDF model, road pBRDF model and snow BRDF model;

[0073] Snow albedo is a crucial parameter in studying local and global energy budgets and climate change. Remote sensing inversion provides a convenient means for obtaining snow albedo. Snow albedo depends primarily on the physical properties of the snow and weather conditions. Existing remote sensing albedo inversions are typically based on the bidirectional reflectance model (BRDF). Snow BRDF models are often derived using snow radiative transfer models.

[0074] Research has shown that the weak absorption properties of snow are primarily influenced by factors such as snow particle size, ice crystal shape, and contaminants. Particle size primarily affects near-infrared absorption, while contaminants primarily affect visible light absorption. The snow BRDF model algorithm used in this example first inverts the snow albedo based on MODIS surface reflectance data, verifies the inversion results, and finally analyzes the reliability of the MODIS snow albedo inversion.

[0075] Currently, computational simulation models are primarily used to analyze the BRDF reflective properties of snow. Computer simulation models are a practical processing algorithm, primarily represented by kernel-driven models. The Monte Carlo method is the theoretical core of this approach, primarily using computers to simulate certain distribution functions. It is a mathematical method with statistical characteristics. The kernel-driven model uses three kernels to describe three types of scattering. These three kernels are weighted to produce the kernel-driven BRDF model expression shown below:

[0076] ;

[0077] in are the incident zenith angle, incident azimuth angle, scattering zenith angle and scattering azimuth angle respectively; Indicates band The BRDF value corresponding to the surface under different ground object coverage; is the geometric optics kernel; the volume scattering kernel is ; 、 、 are the weights of isotropic scattering, geometric target scattering and volume scattering respectively;

[0078] ;

[0079] ;

[0080] in, ；

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] ;

[0086] because 、 、 The three parameters are unknown, so they are obtained through MODIS satellite.

[0087] Where: is the scattering phase angle, represents the incident zenith angle of the microstructure, represents the microstructure reflection zenith angle, represents the incident angle of the micro surface, D represents the geometric occlusion factor, b l It represents the radius of the major axis of the ellipsoid. l has no special definition and is only used to distinguish it from b. r represents the radius of the minor axis of the ellipsoid. t is the intersection of the two ellipses. h represents the relative height. Relative azimuth, is the wavelength, Represents the overlap function between the observed shadow and the light shadow.

[0088] In this embodiment, a five-parameter semi-empirical model is specifically used for the snow BRDF model. The formula is as follows:

[0089] ;

[0090] In the formula, it is composed of the specular reflection component and diffuse reflection component of the sample surface BRDF; is the normal distribution function of the small facets on the sample surface, is the Fresnel approximation function, is the masking function; Indicates the BRDF value of the snow model;

[0091] In this embodiment 、 、 , b are actually four parameters to be determined. and They are used to reflect the magnitude of the specular reflection and diffuse reflection components, respectively, and are related to the roughness and reflectivity of the target surface; b is the parameter reflecting the Fresnel reflection coefficient of the target surface and is related to the refractive index of the target. It represents the angle between the ray from the point on the aperture to the observation point and the normal of the aperture; Represents the angle between the target surface normal and the microfacet normal.

[0092] Sandy land is land with a sandy surface and essentially no vegetation. The surface of the sandy land is flat and rough, and the sand grains therein have uniform and regular shapes. Therefore, this embodiment considers the polarization reflection model of the rough surface when measuring sandy land.

[0093] The sand pBRDF model expression is:

[0094] ;

[0095] ;

[0096] in, is the sand mirror reflection component, is the normal distribution function, θ represents the angle between the incident light and the reflected light; is the masking factor, is the degree of polarization, is a 4×4 Fresnel reflection-Mueller matrix; A is the fitting parameter, β represents the angle between the reflection direction and the normal direction, B is the fitting parameter, represents the refractive index, represents the reflected S wave, represents the reflected P wave, Indicates the BRDF value of the sand model.

[0097] Each blade of grass in a grassland has irregular shapes and shields each other. At the same time, the overall appearance is relatively uniform when observed from a relatively far angle. Therefore, in this embodiment, a six-parameter pBRDF model is used for the grassland.

[0098] The grass pBRDF model expression is:

[0099] ;

[0100] in, represents the slope variance of the rough surface; is the specular reflection coefficient; is the reflectivity of the surface; is the free scattering coefficient; represents the free fitting parameter, represents a six-parameter model; represents the depolarization matrix for volume scattering, represents the backscattered depolarization matrix, is the fitting factor.

[0101] The grass pBRDF model is decomposed into three parts, the first of which is:

[0102] ;

[0103] It is the specular reflection term of the grass background, which represents the reflection amplitude of the entire grass in the specular reflection direction of the incident light source.

[0104] Item 2:

[0105] ;

[0106] is the volume scattering term of the grass background, which represents the reflection relationship between grass plants.

[0107] Item 3:

[0108] ;

[0109] is the backscatter term, which represents the hot spot effect of natural objects and cannot be ignored in this practical application.

[0110] Compared to the grassland background, the model in the woodland background has denser trees, with irregular growth patterns and branches obscuring each other, representing naturally growing ground objects. Compared to the grassland background, the distribution of trees in the woodland is sparser, with more obstruction between branches.

[0111] Compared with the model under grassland background, the polarized BRDF model under forest background has the same specular reflection term and backscattering term, but the volume scattering term changes.

[0112] The forestland pBRDF model expression is:

[0113] = ；

[0114] ;

[0115] ;

[0116] Where: represents the forestland pBRDF model, represents the diffuse component of the forest floor pBRDF model, is the surface autocorrelation length function, Represents the BRDF value of the specular reflection component.

[0117] Roads have both similar and different characteristics compared to sandy terrain. The similarities are that cement and asphalt roads are smooth and have high particle size proximity, so their specular and volumetric reflection terms remain unchanged compared to sandy terrain. The difference is that roads are man-made, not natural, features, but their backscattering can be neglected because of their weak hotspot effects.

[0118] The highway pBRDF model expression is:

[0119] .

[0120] Related results are shown below:

[0121] Background scene modeling and BRDF simulation:

[0122] The various complex backgrounds include: woodlands, deserts, grasslands, roads and snow.

[0123] Simulation modeling is performed based on the materials, characteristics, and roughness of five backgrounds: grassland, forest, snow, desert, and road. Background modeling and corresponding BRDF polar coordinates are as follows: Figure 1-10 As shown:

[0124] During the simulation process, a variety of modeling models were actually applied, including Figure 11 The jeep model is used as an example. The jeep model uses meshing to depict various components and details. Since there is only one target in the following example, the jeep is also a specific target. The total number of meshes in the simplified jeep model is 400,000. The meshing diagram of the jeep model is as follows: Figure 12 As shown:

[0125] The target is coupled to the background and simulated with the BRDF:

[0126] like Figure 13 This is a simulation image of the target under the forest background. Figure 14 This is a schematic diagram of the target simulation results before camouflage. Figure 15 This is a schematic diagram of the target simulation results after camouflage. Figure 16 Spectral reflectance curves of target and grass.

[0127] When the zenith angle and azimuth angle are set to 45° and 10° respectively, the simulation results of 500nm wavelength are as follows Figure 17 As shown; Figure 17 Schematic diagram of target simulation results against a grass background.

[0128] Target recognition algorithm:

[0129] In the target recognition algorithm of this embodiment, three recognition algorithms are constructed according to different application scenarios:

[0130] 1. Object recognition algorithm based on sobel edge detection:

[0131] In object recognition algorithms based on Sobel edge detection, the Sobel operator detects edges based on the phenomenon that the weighted grayscale difference between the upper and lower, and left and right neighbors of a pixel reaches an extreme value at the edge. This smooths noise and provides relatively accurate edge direction information, which is used to calculate image gradients and identify edges and structures within the image. Furthermore, object recognition algorithms based on Sobel edge detection effectively extract edge information from images by performing convolution operations in the spatial domain, making them particularly suitable for detecting horizontal and vertical edges.

[0132] The input image of the Sobel operator is the grayscale image of the measured image. The simulated grayscale image obtained in this embodiment is also a grayscale image. The target edge is the place where the pixel grayscale changes most significantly in the image. The Sobel operator used for edge detection uses the sudden change in the grayscale of the image edge to detect the edge.

[0133] The Sobel operator contains two sets of 3x3 filters, which are sensitive to edges in the horizontal and vertical directions respectively. By using these two operators to perform convolution operations on the input image, the convolution results of each point in the x and y axis directions are obtained: the vertical gradient Gx and the horizontal gradient Gy. A further square sum operation is performed: if the gradient G of a point (x, y) is greater than the preset threshold, then the point (x, y) is considered to be an edge point. However, the edge detection calculation direction based on the Sobel operator is single and it is ineffective for complex textures. Directly using the threshold to judge the edge point will cause a large number of noise points to be misjudged. Therefore, in this embodiment, the interference noise is removed by the four-time morphological corrosion algorithm, and finally an external rectangular frame is generated to obtain the target recognition frame.

[0134] 2. SIFT-based target recognition algorithm:

[0135] Specifically, the scale-invariant feature transform, STFT for short, is used, which includes the following steps:

[0136] First, the algorithm searches for locations in the measured image at all scales. It then uses Gaussian derivatives to identify potential points of interest that are invariant to scale and rotation. Then, it fits the candidate locations to the results of a detailed simulation model to determine their location and scale. Based on the image gradient, these points are assigned one or more directions. Finally, the local gradient of the image is measured at the selected scale in the neighborhood surrounding each key point to describe the key point. Similar to the object recognition algorithm based on Sobel edge detection, this algorithm also identifies the outline of a specific target by obtaining image gradients, then performs morphological erosion to reduce noise, ultimately yielding a specific target identification box.

[0137] 3. Target recognition algorithm based on grayscale threshold segmentation:

[0138] Build a simulation scene, simulate the specific target in the corresponding scene, calculate the BRDF value of the specific target and the background, generate the threshold parameter according to the BRDF value and observation direction of the specific target and background, and then use the threshold parameter to filter out the background in the real image, and then denoise the remaining background to finally generate the target recognition frame.

[0139] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will readily appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A method for detecting and identifying targets in complex backgrounds, characterized in that: The following steps are involved: Determine the application scenario and obtain scenario simulation parameters; Use 3D modeling software to process scene simulation parameters and construct a three-dimensional simulation scene; Use pre-built background models to simulate the spectral characteristics, scattering intensity, and polarization characteristics of specific targets and various complex backgrounds; Performing scene simulation according to scene simulation parameters, three-dimensional simulation scene, and background model, and further acquiring a simulated grayscale image; Compare the grayscale of the simulated grayscale image with the measured image of the application scene, and use the target recognition algorithm to identify and detect specific targets that are uncertain whether they exist in the application scene, and obtain the simulation scene and target recognition results; A variety of complex backgrounds include: woodland, desert, grassland, road and snow; The background model includes the following corresponding to various complex backgrounds: forest pBRDF model, sand pBRDF model, grassland pBRDF model, road pBRDF model and snow BRDF model; Among them, the snow BRDF model expression is: ； formula, Represents the specular reflection component of the sample surface BRDF; is the normal distribution function of the small facets on the sample surface, is the Fresnel approximation function, is the masking function; Indicates the BRDF value of the snow model; 、 、 , b are parameters to be determined; Specifically: and Reflects the magnitude of the specular reflection and diffuse reflection components respectively, and is related to the roughness and reflectivity of the target surface; b is the parameter reflecting the Fresnel reflection coefficient of the target surface and is related to the refractive index of the target. 、 、 、 are the incident zenith angle, incident azimuth angle, scattering zenith angle and scattering azimuth angle respectively; It represents the angle between the ray from the point on the aperture to the observation point and the normal of the aperture; Represents the angle between the target surface normal and the microfacet normal.

2. The method for detecting and identifying targets in a complex background according to claim 1, wherein: The simulation process also includes simulation modeling of the application scene based on the reflectivity of the specific target material itself and the spatial coordinates of the three-dimensional simulation scene.

3. The method for detecting and identifying targets in a complex background according to claim 1, wherein: Acquiring a simulated grayscale image includes the following steps: The sun's incident direction is introduced for spatial calculation, and the full-space BRDF value of a specific target is obtained by traversing the entire space according to the BRDF function; Set the observation direction, calculate the BRDF value of each facet of the specific target material at the corresponding angle, and obtain a simulated grayscale image of the simulated scene.

4. The method for detecting and identifying targets in a complex background according to claim 1, wherein: The sand pBRDF model expression is: ； ； in, is the sand mirror reflection component, is the normal distribution function, It represents the angle between the incident light and the reflected light; Indicates surface roughness; is the masking factor, Indicates relative azimuth; is the degree of polarization, represents the refractive index, is a 4×4 Fresnel reflection-Mueller matrix; is the fitting parameter, Represents the angle between the reflection direction and the normal direction, is the fitting parameter, represents the reflected S wave, represents the reflected P wave, Indicates the BRDF value of the sand model.

5. The method for detecting and identifying targets in a complex background according to claim 4, characterized in that: The grass pBRDF model expression is: ； in, represents the slope variance of the rough surface; is the specular reflection coefficient; is the free scattering coefficient; represents the free fitting parameter, represents a six-parameter model; represents the depolarization matrix for volume scattering, represents the depolarization matrix of backscattering; represents the fitting factor; is the reflectivity of the target surface.

6. The method for detecting and identifying targets in complex backgrounds according to claim 5, characterized in that: The forestland pBRDF model expression is: ； ； ； Where: represents the forestland pBRDF model, represents the diffuse component of the forest floor pBRDF model, is the surface autocorrelation length function, Represents the BRDF value of the specular reflection component.

7. The method for detecting and identifying targets in complex backgrounds according to claim 6, wherein: The highway pBRDF model expression is: 。

Images