A method and system for ecological plot monitoring based on computer vision

Abstract

The present invention is applicable to the technical field of plot monitoring, and provides an ecological plot monitoring method and system based on computer vision, including the following steps: collecting aerial images of an ecological plot, where each aerial image corresponds to an aerial parameter information; performing edge cropping on the aerial images according to the aerial parameter information, and stitching all the edge-cropped aerial images to obtain an ecological plot image; performing image processing and image segmentation on the ecological plot image, segmenting the ecological plot image into several regions, and extracting key feature information of each region; identifying the ecological information of each region according to the key feature information, and at the same time obtaining the recognition credibility; determining the area range and collection coordinates of the regions that need to be recollected based on the recognition credibility; and re-determining the ecological information of each region according to the recollected aerial images. The present invention avoids blind re-sampling according to the recognition credibility, and further improves the accuracy and integrity of the ecological information.

Description

Technical Field

[0001] The present invention relates to the technical field of plot monitoring, and specifically relates to an ecological plot monitoring method and system based on computer vision. Background Art

[0002] With the rapid development of remote sensing technology and computer vision technology, using unmanned aerial vehicle (UAV) aerial photography combined with image processing technology for ecological plot monitoring has become an emerging and efficient method. However, ecological plots usually contain various complex ecological environmental elements, such as ecological types, vegetation types, etc. The performance characteristics of these elements in images are different. How to accurately and efficiently segment images and extract key feature information has become a difficult problem. In addition, due to the influence of factors such as lighting conditions and shooting angles, the images obtained by aerial photography may have problems of incomplete information or distortion, affecting the recognition accuracy of ecological information. Therefore, it is necessary to provide an ecological plot monitoring method and system based on computer vision to solve the above problems. Summary of the Invention

[0003] Aiming at the deficiencies of the existing technology, the purpose of the present invention is to provide an ecological plot monitoring method and system based on computer vision to solve the problems in the above background art.

[0004] The present invention is implemented as follows. An ecological plot monitoring method based on computer vision, the method includes the following steps:

[0005] Collect aerial images of ecological plots, each aerial image corresponds to an aerial parameter information, and the aerial parameter information includes aerial coordinates, aerial height, and shooting focal length;

[0006] Perform edge cropping on the aerial images according to the aerial parameter information, and splice all the edge-cropped aerial images to obtain an ecological plot image;

[0007] Perform image processing and image segmentation on the ecological plot image, segment the ecological plot image into several regions, and extract the key feature information of each region;

[0008] Identify the ecological information of each region according to the key feature information, and at the same time obtain the recognition credibility;

[0009] Determine the area range and collection coordinates that need to be recollected based on the recognition credibility;

[0010] Receive the aerial images of the secondary collection, and re-determine the ecological information of each region according to the aerial images of the secondary collection.

[0011] As a further solution of the present invention: the step of performing edge cropping on the aerial images according to the aerial parameter information and stitching all the edge-cropped aerial images to obtain the ecological plot image specifically includes:

[0012] Calculate the actual ground size L represented by each pixel in the aerial image according to the aerial parameter information. The actual ground size L = 2×h×tanα×W, where h represents the aerial height, W represents the size of the aerial image, and α represents half of the camera view angle, which is calculated through the shooting focal length and the sensor size;

[0013] Determine the edge overlapping area of each aerial image according to the actual ground size and the aerial coordinates, and crop the edge overlapping area;

[0014] Perform image fusion stitching based on feature point matching and the RANSAC algorithm to obtain the ecological plot image.

[0015] As a further solution of the present invention: the step of performing image processing and image segmentation on the ecological plot image, dividing the ecological plot image into several regions, and extracting the key feature information of each region specifically includes:

[0016] Convert the ecological plot image from the RGB color space to the HSV color space, and perform filtering denoising and contrast enhancement;

[0017] Based on the H component of the HSV color space, perform preliminary segmentation on the ecological plot image, and further segment the ecological plot image according to the region growing algorithm;

[0018] Extract the texture features, color features, and shape features of each region to obtain the key feature information.

[0019] As a further solution of the present invention: the step of identifying the ecological information of each region according to the key feature information and obtaining the recognition credibility specifically includes:

[0020] Perform feature fusion according to the texture features, color features, and shape features to obtain a fusion feature vector, and obtain the ecological label of the region based on the fusion feature vector and the trained recognition model;

[0021] Retrieve the reference set corresponding to the ecological label. In the reference set, use the cosine similarity to find the K samples closest to the fusion feature vector;

[0022] Obtain a comprehensive similarity index according to the similarity between the K samples of the nearest neighbor and the fusion feature vector, and determine the recognition credibility according to the comprehensive similarity index.

[0023] As a further solution of the present invention: the step of determining the area range and collection coordinates that need to be recollected based on the recognition credibility specifically includes:

[0024] Determine the area that needs to be recollected based on the recognition credibility, and summarize to obtain the area range;

[0025] Cover the area range based on the maximum secondary acquisition block, and the maximum secondary acquisition block has a constant coverage area;

[0026] Determine the central coordinates of each maximum secondary acquisition block participating in the coverage, and the central coordinates are acquisition coordinates.

[0027] As a further solution of the present invention: the step of preliminarily segmenting the ecological sample plot image based on the H component of the HSV color space and further segmenting the ecological sample plot image according to the region growing algorithm specifically includes:

[0028] Traverse each pixel of the ecological sample plot image, determine the threshold range into which the H component value of each pixel falls, each threshold range corresponds to an ecological element, and perform preliminary segmentation based on the ecological element;

[0029] Select seed points and determine the similarity criterion;

[0030] Starting from each seed point, check whether the pixels in its neighborhood meet the similarity criterion. If so, add the pixel to the current region;

[0031] Repeat the above step until no new pixels can be added to the current region and the seed points complete region growth.

[0032] Another object of the present invention is to provide an ecological sample plot monitoring system based on computer vision, and the system includes:

[0033] An aerial image acquisition module for acquiring aerial images of the ecological sample plot. Each aerial image corresponds to an aerial parameter information, and the aerial parameter information includes aerial coordinates, aerial height, and shooting focal length;

[0034] An image cropping and stitching module for edge cropping the aerial images according to the aerial parameter information, and stitching all the edge-cropped aerial images to obtain an ecological sample plot image;

[0035] A key feature extraction module for performing image processing and image segmentation on the ecological sample plot image, segmenting the ecological sample plot image into several regions, and extracting the key feature information of each region;

[0036] An image region recognition module for identifying the ecological information of each region according to the key feature information and obtaining the recognition credibility at the same time;

[0037] A secondary acquisition determination module for determining the area range and acquisition coordinates that need to be recollected based on the recognition credibility;

[0038] An ecological information determination module, configured to receive the aerial images collected for the second time, and re-determine the ecological information of each area according to the aerial images collected for the second time.

[0039] As a further solution of the present invention: the image cropping and stitching module includes:

[0040] An actual ground size unit, configured to calculate the actual ground size L represented by each pixel in the aerial image according to the aerial photography parameter information, where the actual ground size L = 2×h×tanα×W, where h represents the aerial photography height, W represents the size of the aerial image, and α represents half of the camera viewing angle, which is calculated through the shooting focal length and the sensor size;

[0041] An overlapping area cropping unit, configured to determine the edge overlapping area of each aerial image according to the actual ground size and the aerial photography coordinates, and crop the edge overlapping area;

[0042] An image fusion and stitching unit, configured to perform image fusion and stitching based on feature point matching and the RANSAC algorithm to obtain an ecological sample plot image.

[0043] As a further solution of the present invention: the key feature extraction module includes:

[0044] A color space conversion unit, configured to convert the ecological sample plot image from the RGB color space to the HSV color space for filtering denoising and contrast enhancement;

[0045] A sample plot image segmentation unit, configured to perform preliminary segmentation on the ecological sample plot image based on the H component of the HSV color space, and further segment the ecological sample plot image according to the region growing algorithm;

[0046] A key feature extraction unit, configured to extract the texture feature, color feature, and shape feature of each area to obtain key feature information.

[0047] As a further solution of the present invention: the image area recognition module includes:

[0048] An ecological label recognition unit, configured to perform feature fusion according to the texture feature, color feature, and shape feature to obtain a fusion feature vector, and obtain the ecological label of the area based on the fusion feature vector and the trained recognition model;

[0049] A nearest neighbor sample searching unit, configured to retrieve the reference set corresponding to the ecological label, and in the reference set, use cosine similarity to search for the K nearest neighbor samples to the fusion feature vector;

[0050] An identification credibility unit is used to obtain a comprehensive similarity index based on the similarity between the K nearest neighbor samples and the fusion feature vector, and determine the identification credibility according to the comprehensive similarity index.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] The present invention performs edge cropping on the aerial images according to the aerial photography parameter information, removes the overlapping and blurred parts in the images, reduces the computational amount and interference in subsequent image processing, and improves the processing efficiency and accuracy. The ecological information of each region is identified according to the key feature information, and the identification credibility is obtained at the same time. According to the identification credibility, the range and acquisition coordinates of the region that needs to be recollected are automatically determined, and the region coordinates for secondary acquisition are accurately determined, ensuring the integrity and accuracy of the finally obtained image information and making the feature recognition results more credible. Description of the Drawings

[0053] Figure 1 It is a flowchart of a method for monitoring ecological sample plots based on computer vision.

[0054] Figure 2 It is a flowchart of cropping the aerial images in a method for monitoring ecological sample plots based on computer vision.

[0055] Figure 3 It is a flowchart of image processing in a method for monitoring ecological sample plots based on computer vision.

[0056] Figure 4 It is a flowchart of identifying ecological information in a method for monitoring ecological sample plots based on computer vision.

[0057] Figure 5 It is a flowchart of determining the range and acquisition coordinates of the region to be recollected in a method for monitoring ecological sample plots based on computer vision.

[0058] Figure 6 It is a flowchart of segmenting the ecological sample plot images in a method for monitoring ecological sample plots based on computer vision.

[0059] Figure 7 It is a schematic structural diagram of a system for monitoring ecological sample plots based on computer vision. Detailed Embodiments

[0060] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0061] The following describes the specific implementation of the present invention in detail with reference to specific embodiments.

[0062] As Figure 1 shown, an embodiment of the present invention provides a method for monitoring ecological plots based on computer vision, and the method includes the following steps:

[0063] S100, collect aerial images of ecological plots, each aerial image corresponding to an aerial parameter information, where the aerial parameter information includes aerial coordinates, aerial height, and shooting focal length;

[0064] S200, perform edge cropping on the aerial images according to the aerial parameter information, and splice all the edge-cropped aerial images to obtain an ecological plot image;

[0065] S300, perform image processing and image segmentation on the ecological plot image, segment the ecological plot image into several regions, and extract key feature information of each region;

[0066] S400, identify the ecological information of each region according to the key feature information, and obtain the recognition credibility at the same time;

[0067] S500, determine the area range and collection coordinates that need to be recollected based on the recognition credibility;

[0068] S600, receive the aerial images collected for the second time, and re-determine the ecological information of each region according to the aerial images collected for the second time.

[0069] In the embodiments of the present invention, when collecting aerial images of ecological sample plots, an aerial parameter information will be automatically set for each aerial image. The aerial parameter information includes aerial coordinates, aerial height, and shooting focal length. A positioning system and a height sensor are installed on the unmanned aerial vehicle, which can obtain coordinates and height in real time. Then, the embodiments of the present invention will perform edge cropping on the aerial images according to the aerial parameter information, and splice all the edge-cropped aerial images to obtain an ecological sample plot image. In this way, the fusion splicing effect is better. Then, image processing and image segmentation will be performed on the ecological sample plot image, and the ecological sample plot image will be segmented into several regions. The ecological type of each region is the same, and the ecological types include grassland, forest, water body, farmland, etc. The key feature information of each region will be extracted, and then the ecological information of each region will be identified according to the key feature information. The ecological information is a refinement of the ecological type. For example, the ecological information of forest ecology is the specific vegetation type. At the same time, the recognition credibility is obtained. The recognition credibility reflects the accuracy of the recognized ecological information. If the recognition credibility is low, it means that the recognition result is not accurate enough. The embodiments of the present invention will determine the area range and collection coordinates that need to be recollected according to all the recognition credibilities, perform secondary aerial photography collection based on the area range and collection coordinates, obtain the aerial images of the secondary collection, and re-determine the ecological information of each region according to the aerial images of the secondary collection. In this way, it is ensured that the finally obtained image information is complete and accurate, and the feature recognition result is credible.

[0070] As Figure 2 shown, as a preferred embodiment of the present invention, the step of performing edge cropping on the aerial image according to the aerial parameter information and splicing all the edge-cropped aerial images to obtain an ecological sample plot image specifically includes:

[0071] S201, calculate the actual ground size L represented by each pixel in the aerial image according to the aerial parameter information. The actual ground size L = 2×h×tanα×W, where h represents the aerial height, W represents the aerial image size, and α represents half of the camera view angle, which is calculated through the shooting focal length and the sensor size;

[0072] S202, determine the edge overlapping area of each aerial image according to the actual ground size and the aerial coordinates, and crop the edge overlapping area;

[0073] S203, perform image fusion splicing based on feature point matching and the RANSAC algorithm to obtain an ecological sample plot image.

[0074] In the embodiments of the present invention, in order to perform precise cutting and splicing, first, the actual ground size L represented by each pixel in the aerial image needs to be calculated according to the aerial photography parameter information. The actual ground size L = 2×h×tanα×W. Then, the edge overlapping area of each aerial image is determined according to the actual ground size and the aerial photography coordinates, and the edge overlapping area is cut. The aerial photography coordinates are the center points of each aerial image. Cutting the edge area can, on the one hand, remove the overlapping part, and on the other hand, remove the partially blurred image at the edge. Finally, image fusion and splicing are performed according to feature point matching (such as SIFT, SURF) and the RANSAC algorithm to obtain the ecological sample plot image.

[0075] As Figure 3 shown, as a preferred embodiment of the present invention, the step of performing image processing and image segmentation on the ecological sample plot image, and segmenting the ecological sample plot image into several regions, and extracting the key feature information of each region specifically includes:

[0076] S301, converting the ecological sample plot image from the RGB color space to the HSV color space, and performing filtering to remove noise and contrast enhancement;

[0077] S302, performing preliminary segmentation on the ecological sample plot image based on the H component of the HSV color space, and further segmenting the ecological sample plot image according to the region growing algorithm;

[0078] S303, extracting the texture feature, color feature, and shape feature of each region to obtain the key feature information.

[0079] In the embodiments of the present invention, the ecological sample plot image will be converted from the RGB color space to the HSV color space. The H (hue) component in the HSV color space is particularly useful for distinguishing different ecological types. Then, algorithms such as Gaussian filtering and median filtering are applied to remove the noise in the image and perform contrast enhancement. Then, preliminary segmentation is performed on the ecological sample plot image based on the H component of the HSV color space, and further segmentation is performed on the ecological sample plot image according to the region growing algorithm to obtain several regions. Finally, the texture feature, color feature, and shape feature of each region are extracted to obtain the key feature information, where the texture feature is extracted using methods such as gray-level co-occurrence matrix (GLCM) and local binary pattern (LBP); the color feature is obtained by calculating the average color, color histogram, etc. of each region.

[0080] As Figure 4 shown, as a preferred embodiment of the present invention, the step of identifying the ecological information of each region according to the key feature information and obtaining the recognition credibility specifically includes:

[0081] S401. Perform feature fusion based on texture features, color features, and shape features to obtain a fused feature vector, and obtain the ecological label of the region based on the fused feature vector and the trained recognition model.

[0082] S402. Retrieve the reference set corresponding to the ecological label. In the reference set, use cosine similarity to find the K samples closest to the fused feature vector.

[0083] S403. Obtain a comprehensive similarity index based on the similarity between the K closest samples and the fused feature vector, and determine the recognition credibility according to the comprehensive similarity index.

[0084] In the embodiment of the present invention, a recognition model is pre-trained. Specifically, models such as decision trees, random forests, support vector machines (SVMs), and neural networks are selected for training. The regional data with known ecological information (i.e., labeled data) is used to train the recognition model. During or after the training process, a validation set or a test set is used to evaluate the performance of the model. When in use, feature fusion is performed based on texture features, color features, and shape features to obtain a fused feature vector, and the ecological label of the region is obtained based on the fused feature vector and the trained recognition model; the reference set corresponding to the ecological label is retrieved. The training set used during training and the validation set used during validation are both reference sets. In the reference set, cosine similarity is used to find the K samples closest to the fused feature vector. K is a fixed value set in advance. Then, a comprehensive similarity index is obtained based on the similarity between the K closest samples and the fused feature vector. The comprehensive similarity index can be the average similarity of the K samples. Finally, the recognition credibility is determined according to the comprehensive similarity index, and the recognition credibility is proportional to the comprehensive similarity index.

[0085] As Figure 5 shown, as a preferred embodiment of the present invention, the step of determining the area range and acquisition coordinates that need to be recollected based on the recognition credibility specifically includes:

[0086] S501. Determine the areas that need to be recollected based on the recognition credibility, and summarize to obtain the area range.

[0087] S502. Cover the area range based on the maximum secondary acquisition block, and the maximum secondary acquisition block has a constant coverage area.

[0088] S503. Determine the center coordinates of each maximum secondary acquisition block participating in the coverage, and the center coordinates are the acquisition coordinates.

[0089] In the embodiments of the present invention, an area that needs to be recollected is determined based on the recognition credibility. For example, when the recognition credibility is lower than the credibility threshold, the corresponding area image needs to be recollected. The area ranges of these areas are summarized, and then the area range is covered based on the maximum secondary collection block. The maximum secondary collection block has a constant coverage area. The maximum secondary collection block is the block size obtained using the set aerial photography height and shooting focal length, and the content in the block can be clearly photographed. The principle of covering the area range is to use as few maximum secondary collection blocks as possible to comprehensively cover the area range. Finally, the central coordinates of each maximum secondary collection block participating in the coverage are determined, and the central coordinates are the collection coordinates.

[0090] As Figure 6 shown, as a preferred embodiment of the present invention, the step of preliminarily segmenting the ecological sample plot image based on the H component in the HSV color space and further segmenting the ecological sample plot image according to the region growing algorithm specifically includes:

[0091] S3021, traverse each pixel of the ecological sample plot image, determine the threshold range into which the H component value of each pixel falls, each threshold range corresponds to an ecological element, and perform preliminary segmentation based on the ecological element;

[0092] S3022, select seed points and determine the similarity criterion;

[0093] S3023, starting from each seed point, check whether the pixels in its neighborhood satisfy the similarity criterion. If so, add the pixel to the current region;

[0094] S3024, repeat the above step until no new pixels can be added to the current region and the seed points complete region growing.

[0095] In the embodiments of the present invention, when performing image segmentation, first, each pixel of the ecological sample plot image is traversed to determine the threshold range into which the H component value of each pixel falls. Each threshold range corresponds to an ecological element, and the ecological elements are grassland, forest, water body, farmland, etc. Then, preliminary segmentation is performed based on the ecological element type. Next, seed points need to be selected. The seed points can be randomly selected by setting the seed density, and then the similarity criterion is determined. The similarity criterion can be determined based on the H component, S component, and V component. Starting from each seed point, check whether the pixels in its neighborhood satisfy the similarity criterion. If so, add the pixel to the current region. Repeat this step until no new pixels can be added to the current region and the seed points complete region growing. Finally, adjacent similar regions need to be merged or too small noise regions need to be removed.

[0096] As Figure 7As shown in the figure, an embodiment of the present invention further provides an ecological plot monitoring system based on computer vision, and the system includes:

[0097] An aerial image acquisition module 100, configured to acquire aerial images of ecological plots. Each aerial image corresponds to an aerial parameter information, and the aerial parameter information includes aerial coordinates, aerial height, and shooting focal length;

[0098] An image cropping and stitching module 200, configured to perform edge cropping on the aerial images according to the aerial parameter information, and stitch all the edge-cropped aerial images to obtain an ecological plot image;

[0099] A key feature extraction module 300, configured to perform image processing and image segmentation on the ecological plot image, segment the ecological plot image into several regions, and extract the key feature information of each region;

[0100] An image region recognition module 400, configured to recognize the ecological information of each region according to the key feature information, and obtain the recognition credibility at the same time;

[0101] A secondary acquisition determination module 500, configured to determine the region range and acquisition coordinates that need to be re-acquired based on the recognition credibility;

[0102] An ecological information determination module 600, configured to receive the aerial images acquired secondarily, and re-determine the ecological information of each region according to the aerial images acquired secondarily.

[0103] As a preferred embodiment of the present invention, the image cropping and stitching module 200 includes:

[0104] An actual ground size unit, configured to calculate the actual ground size L represented by each pixel in the aerial image according to the aerial parameter information. The actual ground size L = 2×h×tanα×W, where h represents the aerial height, W represents the aerial image size, and α represents half of the camera view angle, which is calculated through the shooting focal length and the sensor size;

[0105] An overlapping region cropping unit, configured to determine the edge overlapping regions of each aerial image according to the actual ground size and the aerial coordinates, and crop the edge overlapping regions;

[0106] An image fusion and stitching unit, configured to perform image fusion and stitching based on feature point matching and the RANSAC algorithm to obtain an ecological plot image.

[0107] As a preferred embodiment of the present invention, the key feature extraction module 300 includes:

[0108] A color space conversion unit, configured to convert the ecological plot image from the RGB color space to the HSV color space for filtering and denoising and contrast enhancement;

[0109] The plot image segmentation unit is used to perform preliminary segmentation on the ecological plot image based on the H component of the HSV color space, and further segment the ecological plot image according to the region growing algorithm;

[0110] The key feature extraction unit is used to extract the texture feature, color feature and shape feature of each region to obtain key feature information.

[0111] As a preferred embodiment of the present invention, the image region recognition module 400 includes:

[0112] The ecological label recognition unit is used to perform feature fusion based on the texture feature, color feature and shape feature to obtain a fusion feature vector, and obtain the ecological label of the region based on the fusion feature vector and the trained recognition model;

[0113] The nearest neighbor sample searching unit is used to retrieve the reference set corresponding to the ecological label, and in the reference set, use cosine similarity to find the K nearest neighbor samples to the fusion feature vector;

[0114] The recognition credibility unit is used to obtain a comprehensive similarity index according to the similarity between the K nearest neighbor samples and the fusion feature vector, and determine the recognition credibility according to the comprehensive similarity index.

[0115] The above only describes the preferred embodiments of the present invention in detail, and does not limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0116] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps does not have a strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least a part of other steps or sub-steps or stages of other steps.

[0117] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0118] After considering the specification and the disclosure of the embodiments, those skilled in the art will readily conceive of other embodiments of the present disclosure. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include known common knowledge or conventional technical means in the technical field not disclosed in the present disclosure. The specification and the embodiments are only regarded as exemplary, and the true scope and spirit of the present disclosure are pointed out by the claims.

Claims

1. A computer vision-based ecological plot monitoring method, characterized in that: The method comprises the following steps: Collecting aerial images of the ecological sample plot, each aerial image corresponds to an aerial parameter information, and the aerial parameter information includes aerial coordinates, aerial altitude and shooting focal length; The aerial images are edge-cropped according to the aerial photography parameter information, and all the aerial images after edge-cropping are spliced ​​to obtain the ecological sample plot image; The ecological sample plot image is processed and segmented, and the ecological sample plot image is segmented into several regions, and the key feature information of each region is extracted; specifically, the ecological sample plot image is converted from the RGB color space to the HSV color space, and filtering denoising and contrast enhancement are performed; the ecological sample plot image is preliminarily segmented based on the H component of the HSV color space, and the ecological sample plot image is further segmented according to the regional growing algorithm; the texture features, color features and shape features of each region are extracted to obtain the key feature information; Identify the ecological information of each area according to the key feature information, and obtain the recognition credibility at the same time; specifically include: perform feature fusion according to texture features, color features and shape features to obtain a fused feature vector, and obtain the ecological label of the area based on the fused feature vector and the trained recognition model; retrieve the reference set corresponding to the ecological label, and use the cosine similarity to find the K samples that are the nearest neighbors of the fused feature vector in the reference set; obtain a comprehensive similarity index according to the similarity between the K nearest neighbor samples and the fused feature vector, and determine the recognition credibility according to the comprehensive similarity index; Determine the area range and acquisition coordinates that need to be re-collected based on the recognition credibility; specifically include: determine the area that needs to be re-collected based on the recognition credibility, and summarize the area range; cover the area range based on the maximum secondary acquisition block, and the maximum secondary acquisition block has a constant coverage area; determine the center coordinates of each maximum secondary acquisition block involved in the coverage, and the center coordinates are the acquisition coordinates; wherein the maximum secondary acquisition block is a block size obtained by using the set aerial photography height and shooting focal length, and the content in the shooting block can be clearly photographed. The principle of covering the area range is: use the minimum maximum secondary acquisition block to cover the area range in all directions; The secondary collected aerial images are received, and the ecological information of each area is re-determined based on the secondary collected aerial images.

2. The ecological plot monitoring method based on computer vision according to claim 1, characterized in that: The step of performing edge cropping on the aerial image according to the aerial photography parameter information and splicing all the edge cropped aerial images to obtain the ecological sample plot image specifically includes: The actual ground size L represented by each pixel in the aerial image is calculated based on the aerial photography parameter information. The actual ground size L=2×h×tanα×W, where h represents the aerial photography altitude, W represents the aerial image size, and α represents half of the camera's viewing angle, which is calculated based on the shooting focal length and sensor size. Determine the edge overlap area of ​​each aerial image according to the actual ground size and the aerial coordinates, and crop the edge overlap area; Image fusion and splicing are performed based on feature point matching and RANSAC algorithm to obtain ecological sample plot images.

3. The ecological plot monitoring method based on computer vision according to claim 1, characterized in that: The steps of preliminarily segmenting the ecological sample plot image based on the H component of the HSV color space and further segmenting the ecological sample plot image according to the region growing algorithm specifically include: Traverse each pixel of the ecological sample plot image, determine the threshold range that the H component value of each pixel falls into, each threshold range corresponds to an ecological element, and perform preliminary segmentation based on the ecological element; Select seed points and determine similarity criteria; Starting from each seed point, check whether the pixels in its neighborhood meet the similarity criterion. If so, add the pixels to the current area; Repeat the previous step until no new pixels can be added to the current region and the seed point completes the region growth.

4. An ecological plot monitoring system based on computer vision, characterized in that: The system comprises: An aerial image acquisition module is used to collect aerial images of ecological plots, each aerial image corresponds to an aerial parameter information, and the aerial parameter information includes aerial coordinates, aerial altitude and shooting focal length; An image cropping and splicing module is used to crop the edges of the aerial images according to the aerial photography parameter information, and to splice all the aerial images after the edges have been cropped to obtain an ecological sample plot image; The key feature extraction module is used to process and segment the ecological sample image, segment the ecological sample image into several regions, and extract the key feature information of each region; specifically, it includes: converting the ecological sample image from RGB color space to HSV color space, filtering denoising and contrast enhancement; performing preliminary segmentation of the ecological sample image based on the H component of the HSV color space, and further segmenting the ecological sample image according to the regional growing algorithm; extracting the texture features, color features and shape features of each region to obtain the key feature information; The image region recognition module is used to identify the ecological information of each region based on the key feature information and obtain the recognition credibility at the same time; perform feature fusion based on texture features, color features and shape features to obtain a fused feature vector, and obtain the ecological label of the region based on the fused feature vector and the trained recognition model; retrieve the reference set corresponding to the ecological label, and use the cosine similarity to find the K samples that are the nearest neighbors of the fused feature vector in the reference set; obtain a comprehensive similarity index based on the similarity between the K nearest neighbor samples and the fused feature vector, and determine the recognition credibility based on the comprehensive similarity index; The secondary collection determination module is used to determine the area range and collection coordinates that need to be re-collected based on the recognition credibility; specifically, it includes: determining the area that needs to be re-collected based on the recognition credibility, and summarizing the area range; covering the area range based on the maximum secondary collection block, and the maximum secondary collection block has a constant coverage area; determining the center coordinates of each maximum secondary collection block involved in the coverage, and the center coordinates are the collection coordinates; wherein the maximum secondary collection block is a block size obtained by using the set aerial photography height and shooting focal length, and the content in the shooting block can be clearly photographed. The principle of covering the area range is: using the minimum maximum secondary collection block to cover the area range in all directions; The ecological information determination module is used to receive the secondary collected aerial images and re-determine the ecological information of each area based on the secondary collected aerial images.

5. The computer vision-based ecological plot monitoring system according to claim 4, characterized in that: The image cropping and splicing module includes: The actual ground size unit is used to calculate the actual ground size L represented by each pixel in the aerial image according to the aerial photography parameter information. The actual ground size L=2×h×tanα×W, where h represents the aerial photography altitude, W represents the aerial image size, and α represents half of the camera viewing angle, which is calculated by the shooting focal length and the sensor size. The overlap region clipping unit is used to determine the edge overlap region of each aerial image according to the actual ground size and the aerial coordinates, and to clip the edge overlap region; The image fusion and stitching unit is used to perform image fusion and stitching based on feature point matching and RANSAC algorithm to obtain ecological sample plot images.

Images