Complex jungle environment three-dimensional space reconstruction method based on image recognition

Abstract

The invention provides a complex jungle environment three-dimensional space reconstruction method based on image recognition, and relates to the technical field of three-dimensional reconstruction, and the method comprises the steps: carrying out the arrangement of an environment monitoring network through combining the initial detection information of a target jungle, and obtaining a jungle environment monitoring network; determining M effective image acquisition space sets; determining K image acquisition overlapping space sets; obtaining K overlapping space splicing path sets; a path set is spliced in the determined matching overlapping space; and extracting acquired images of the M image acquisition devices in the current reconstruction window to obtain M real-time acquired images, performing three-dimensional reconstruction based on the M real-time acquired images, splicing the reconstruction space in combination with the matched overlapping space splicing path set, and determining a target reconstruction environment three-dimensional space of the target jungle. The technical problem that in the prior art, the accuracy is low when three-dimensional space reconstruction is carried out in the complex jungle environment is solved, and the effect of improving the three-dimensional reconstruction precision and quality is achieved.

Description

Technical Field

[0001] The present application relates to the field of three-dimensional reconstruction technology, and in particular to a three-dimensional space reconstruction method for a complex jungle environment based on image recognition. Background Art

[0002] Traditional single-view or few-device acquisition methods often cannot fully cover all the details in the jungle, resulting in limited image data, especially in areas with dense vegetation or complex terrain. In addition, due to the large differences in lighting, occlusion, and vegetation height in different areas of the jungle, even if multiple devices are used for simultaneous acquisition, due to improper handling of the overlapping areas of vision between devices, the spatial reconstruction results may contain redundant data or inaccurate overlap, which in turn affects the accuracy of the overall 3D model.

[0003] In summary, the prior art has the technical problem of low accuracy in three-dimensional space reconstruction in a complex jungle environment. Summary of the invention

[0004] The purpose of this application is to provide a method for three-dimensional space reconstruction of a complex jungle environment based on image recognition, so as to solve the technical problem of low accuracy in three-dimensional space reconstruction of a complex jungle environment in the prior art.

[0005] In view of the above problems, the present application provides a method for reconstructing a three-dimensional space of a complex jungle environment based on image recognition, wherein the method comprises:

[0006] Deploy an environmental monitoring network based on the initial detection information of the target jungle to obtain a jungle environmental monitoring network, wherein the jungle environmental monitoring network includes M image acquisition devices, M device deployment positions, and M device deployment angles, where M is an integer greater than or equal to 1;

[0007] Interacting the basic device information of the M image acquisition devices, determining M effective image acquisition space sets of the M image acquisition devices under K monitoring scenes and the M device deployment angles, each effective image acquisition space is an area range of an image acquisition device in a monitoring scene where the image quality meets the requirements, wherein K is an integer greater than or equal to 1;

[0008] Using the K monitoring scenes as indexes, performing overlapping space identification on the M valid image acquisition space sets and the M equipment deployment positions, and determining K image acquisition overlapping space sets, wherein each image acquisition overlapping space includes multiple overlapping objects of interest and multiple superimposed pixel point gradients;

[0009] Traversing the K image acquisition overlapping space sets to perform spatial fusion recognition of the target of interest, determining a spatial splicing path according to the recognition result, and obtaining a set of K overlapping spatial splicing paths;

[0010] Collecting node monitoring scene features of the jungle environment monitoring network in the current reconstruction window, matching them with the K monitoring scenes to obtain a matching monitoring scene, and determining a matching overlapping space splicing path set according to the matching monitoring scene and the K overlapping space splicing path sets;

[0011] The captured images of M image acquisition devices in the current reconstruction window are extracted to obtain M real-time captured images, three-dimensional reconstruction is performed based on the M real-time captured images, and the reconstructed space is spliced ​​in combination with the matching overlapping space splicing path set to determine the three-dimensional space of the target reconstruction environment of the target jungle.

[0012] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0013] The present application obtains a jungle environment monitoring network by combining the initial detection information of the target jungle to deploy an environmental monitoring network, wherein the jungle environment monitoring network includes M image acquisition devices, M device deployment positions and M device deployment angles, where M is an integer greater than or equal to 1, and then interacts with the basic device information of the M image acquisition devices to determine M valid image acquisition space sets of the M image acquisition devices under K monitoring scenes and M device deployment angles, each valid image acquisition space being an area range where the image quality of an image acquisition device meets the requirements in a monitoring scene, where K is an integer greater than or equal to 1, and then uses the K monitoring scenes as indexes to identify overlapping spaces of the M valid image acquisition space sets and the M device deployment positions, and determines K image acquisition overlapping space sets, wherein each image acquisition The overlapping space includes multiple overlapping targets of interest and multiple superimposed pixel gradients, and then traverses K image acquisition overlapping space sets to perform spatial fusion recognition of the target of interest, determines the spatial splicing path according to the recognition result, obtains K overlapping space splicing path sets, collects node monitoring scene features of the jungle environment monitoring network in the current reconstruction window, matches with K monitoring scenes, obtains matching monitoring scenes, and determines matching overlapping space splicing path sets according to matching monitoring scenes and K overlapping space splicing path sets, and then extracts the collected images of M image acquisition devices in the current reconstruction window to obtain M real-time collected images, performs three-dimensional reconstruction based on the M real-time collected images, and splices the reconstruction space in combination with the matching overlapping space splicing path set to determine the target reconstruction environment three-dimensional space of the target jungle. The technical effect of improving the accuracy and quality of three-dimensional reconstruction can be achieved.

[0014] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0016] Figure 1 A schematic diagram of the process of a method for reconstructing three-dimensional space of a complex jungle environment based on image recognition in this application;

[0017] Figure 2 This is a schematic diagram of a process for determining M valid image acquisition space sets in a three-dimensional space reconstruction method for a complex jungle environment based on image recognition in this application. DETAILED DESCRIPTION

[0018] The present application solves the technical problem of low accuracy in three-dimensional space reconstruction of complex jungle environments in the prior art by providing a method for three-dimensional space reconstruction of complex jungle environments based on image recognition.

[0019] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0020] Examples, such as Figure 1 As shown, the present application provides a method for reconstructing a three-dimensional space of a complex jungle environment based on image recognition, and the method specifically comprises the following steps:

[0021] Step S100: deploying an environmental monitoring network in combination with the initial detection information of the target jungle to obtain a jungle environmental monitoring network, wherein the jungle environmental monitoring network includes M image acquisition devices, M device deployment positions and M device deployment angles, where M is an integer greater than or equal to 1;

[0022] In a possible embodiment, the target jungle is any jungle that needs three-dimensional space reconstruction. The initial detection information reflects the overall characteristics of the target jungle, including topography, vegetation density, tree height and other information. The jungle environment monitoring network is used to collect images of the target jungle, thereby providing image data support for subsequent three-dimensional space reconstruction.

[0023] Preferably, a remote sensing device (such as a drone, LIDAR, satellite remote sensing) is used to conduct a preliminary detection of the target jungle to obtain the overall characteristics of the jungle, thereby obtaining the initial detection information. Optionally, the initial detection information includes topography (such as mountains, hills, plains, etc., which mainly affect the location and height of the equipment), vegetation density and distribution (such as tree height, plant density, which determine the field of view of the equipment and the choice of viewing angle), lighting conditions and occlusion (used to understand the light intensity in different areas and the occlusion of vegetation or other obstacles). Based on the initial detection information, a person skilled in the art will analyze and determine the type of image acquisition equipment (such as a high-resolution camera, a multi-spectral camera, or a device with infrared function, etc.), and determine the equipment layout location and equipment layout angle according to the area of ​​the target jungle.

[0024] Preferably, by analyzing the initial detection data, the coverage of the target jungle is calculated and the number of devices to be deployed is determined. The value of M must be sufficient to cover the entire area of ​​the target jungle and ensure that the overlapping area can provide sufficient image data for three-dimensional reconstruction. Exemplarily, the target jungle is divided into several monitoring areas to ensure that at least one image acquisition device covers each area. In flat areas, the equipment should be evenly distributed; in complex terrain such as slopes and valleys, the equipment should be deployed at high or low places to ensure that terrain features at different heights can be captured. In densely vegetated areas, a wider field of view can be obtained by deploying equipment at high positions (such as drones or installed on tree canopies) to avoid blind spots caused by tree canopy obstruction. The layout position of the equipment should ensure that each area can be covered by multiple devices, especially key jungle areas (such as special landforms or densely vegetated areas). By reasonably setting the overlapping field of view between devices, ensure that there is enough image overlap for subsequent splicing. When setting the device layout angle, the horizontal angle of each device should cover the largest surrounding monitoring area. The device can adjust the horizontal angle around its layout point to ensure that enough jungle features are captured in the field of view.

[0025] By combining the initial detection information of the jungle environment, the layout positions and angles of M image acquisition devices are determined to ensure that the devices can effectively cover all important areas in the jungle and have sufficient overlapping fields of view to provide data support for subsequent three-dimensional reconstruction.

[0026] Step S200: interacting with the basic device information of the M image acquisition devices, determining M effective image acquisition space sets of the M image acquisition devices under K monitoring scenes and the M device deployment angles, each effective image acquisition space being an area range where the image quality of an image acquisition device meets the requirements under one monitoring scene, wherein K is an integer greater than or equal to 1;

[0027] Further, such as Figure 2 As shown, the basic device information of the M image acquisition devices is interacted to determine the M effective image acquisition space sets of the M image acquisition devices under the K monitoring scenes and the M device layout angles. The step S200 of the embodiment of the present application also includes:

[0028] Obtaining K jungle environment feature sets of the K monitoring scenes;

[0029] Interacting with the jungle environment monitoring network to obtain M sets of basic information of the M image acquisition devices, and performing test image acquisition in combination with the M device layout angles and the K jungle environment feature sets to obtain M test image sets, wherein each test image set includes K test images, and each test image corresponds to a monitoring scene;

[0030] Traversing the M test image sets to perform pixel point gradient calculations, and determining M image valid area sets according to the calculation results;

[0031] The objects of interest are extracted from the M sets of effective image regions, and spatial coordinate mapping analysis is performed in combination with the extraction results, and the M sets of effective image acquisition spaces are determined according to the analysis results.

[0032] Furthermore, each of the K jungle environment feature sets includes lighting conditions, weather conditions, and monitoring seasons.

[0033] In one embodiment, the basic device information of the M image acquisition devices is obtained to master the image capturing capabilities of the M image acquisition devices, and the spatial ranges of the M image acquisition devices that can capture valid images in different monitoring scenes are determined in combination with the K monitoring scenes and the deployment angles of each device, so as to obtain the M valid image acquisition space sets. Each valid image acquisition space is the area range of an image acquisition device in a monitoring scene where the image quality meets the requirements.

[0034] Optionally, step S200 of the present application gradually determines the effective image acquisition space of each device in different monitoring scenarios through the interaction of basic device information, monitoring scenarios and deployment angles, using pixel gradient calculation and object of interest extraction. This process ensures that the complex environment of the jungle can be fully covered through multi-angle and multi-scene image acquisition, and provides high-quality image data for subsequent three-dimensional space reconstruction.

[0035] In one embodiment, the K monitoring scenes are determined based on the initial detection information of the target jungle, wherein the K monitoring scenes are the environmental conditions with the top K occurrence frequencies in the target jungle. The K jungle environmental feature sets are determined based on the terrain and climate conditions in the initial detection information of the target jungle, wherein each jungle environmental feature set includes lighting conditions, weather conditions, and monitoring seasons.

[0036] For example, the lighting conditions include the light intensity at different time periods (morning, noon, evening). The weather conditions include the impact of different weather conditions such as sunny days, cloudy days, rainy days or foggy days on image acquisition. The monitoring season includes the impact of seasonal changes such as dense vegetation in spring and falling leaves in autumn on the jungle environment.

[0037] Optionally, the jungle environment monitoring network is interacted with to obtain M sets of basic device information of the M image acquisition devices. Each set of basic device information includes device parameters (such as resolution, focal length, field of view, sensor characteristics, etc.), device age and device manufacturer, etc. Under K sets of jungle environment characteristics, image acquisition tests are performed on the M image acquisition devices at the M device layout angles, and the image with the best image quality in the test is selected as the test image to obtain the M test image sets. In order to facilitate subsequent three-dimensional reconstruction, the test images of each image acquisition device in different monitoring scenarios are grouped together.

[0038] Optionally, in order to identify the clear area in the set of M test images, it is necessary to calculate the gradient of the pixels of each test image in the set of M test images. The larger the pixel gradient, the clearer the image at that location, and the smaller the pixel gradient, the blurrier the image at that location. According to the calculation results, the edge of the clear image is identified to obtain the set of effective areas of the M images. Optionally, the Canny edge detection algorithm is used to calculate the pixel gradient in each image, and then the clarity of different areas in the image is evaluated.

[0039] Preferably, the M sets of effective image areas include situations where images are captured at different positions in the target jungle. By extracting the objects of interest in the image, the positions of the objects of interest to be analyzed in the target jungle are determined, thereby determining the spatial range in which the M image acquisition devices can capture effective images, that is, the M effective image acquisition space sets.

[0040] Further, the M test image sets are traversed to perform pixel gradient calculation, and M image valid area sets are determined according to the calculation results. Step S200 of the embodiment of the present application further includes:

[0041] Randomly extracting a first test image from the set of M test images, and performing gradient calculation on pixel points in the first test image to obtain a plurality of first pixel point gradients;

[0042] Taking the pixel points corresponding to the N first pixel point gradients located in the first N positions among the multiple first pixel point gradients as the center points of N first effective sub-regions, taking the center points of the N first effective sub-regions as the starting point, performing effective sub-region diffusion according to a preset area diffusion step length until a preset diffusion stop condition is met, and obtaining N first effective sub-regions, wherein N is an integer greater than or equal to 1, and the preset area diffusion step length is the maximum pixel point gradient difference allowed in a single diffusion;

[0043] Taking a union of the N first valid sub-regions to obtain a first image valid region;

[0044] Pixel point gradient calculation is performed on the M test image sets according to the preset area diffusion step length to determine the M image valid area sets.

[0045] Furthermore, step S200 of the embodiment of the present application further includes:

[0046] Taking the center points of the N first effective sub-regions as starting points, effective sub-region diffusion is performed according to a preset region diffusion step length to obtain N first initial diffusion sub-regions;

[0047] Performing edge diffusion on the N first initial diffusion sub-regions again according to the preset regional diffusion step length to obtain N first-stage diffusion sub-regions;

[0048] Determine whether a difference between a pixel point gradient mean value of the N first stage diffusion sub-regions and a pixel point gradient mean value of the N first initial diffusion sub-regions satisfies a preset difference value, and if so, the preset diffusion stop condition is satisfied, and the N first stage diffusion sub-regions are used as N first effective sub-regions;

[0049] If not, edge diffusion is performed again on the N first-stage diffusion sub-regions according to the preset region diffusion step length until a preset diffusion stop condition is met, thereby obtaining the N first effective sub-regions.

[0050] In one embodiment, an image (referred to as the first test image) is randomly selected from a set of M test images as an initial sample for subsequent gradient calculation and effective area identification. Preferably, the first test image is preprocessed, including denoising and contrast enhancement, to ensure the accuracy of gradient calculation. Further, the Canny edge detection gradient algorithm is used to calculate the gradient value of each pixel in the first test image, and the gradient strength of each pixel is obtained, thereby obtaining the multiple first pixel gradients. The gradient reflects the gradient strength and direction of each pixel in the image.

[0051] Preferably, the pixel points corresponding to the N first pixel point gradients located in the first N positions among the plurality of first pixel point gradients are used as the center points of the N first effective sub-regions. Since the N first pixel point gradients are relatively large, the corresponding first pixel points can be used as the starting point for effective sub-region diffusion, thereby improving the efficiency of effective region screening.

[0052] Taking the center points of the N first effective sub-regions as the starting point, effective sub-region diffusion is performed according to a preset area diffusion step length until a preset diffusion stop condition is met, thereby obtaining N first effective sub-regions, wherein N is an integer greater than or equal to 1, and the preset area diffusion step length is the maximum pixel point gradient difference allowed when performing a single diffusion set by a technician in this field.

[0053] Preferably, the N first effective sub-regions are relatively clear sub-regions formed by diffusion outward with the center points of the N first effective sub-regions as the center. By taking the union of the N first effective sub-regions, a first image effective region with higher quality in the first test image can be obtained. Based on the same principle as obtaining the effective region of the first image, the pixel gradient calculation is performed on the M test image set according to the preset area diffusion step length to determine the M image effective region set. The goal of identifying the effective region of the M test image set is achieved.

[0054] Preferably, first, effective sub-region diffusion is performed with the center points of the N first effective sub-regions as the starting point according to the preset regional diffusion step length to obtain N first initial diffusion sub-regions. That is to say, the N first initial diffusion sub-regions are regions surrounded by multiple first pixel points located at the outermost periphery of the first pixel point set corresponding to multiple first pixel point gradients obtained after the pixel point gradient screening that satisfies the preset regional diffusion step length is performed with the center points of the N first effective sub-regions as the starting point.

[0055] Then, the first pixel points at the outermost part of the first initial diffusion sub-region are used as diffusion starting points, and diffusion is performed according to the preset regional diffusion step length to obtain the N first-stage diffusion sub-regions. The mean value of the pixel gradients in the N first-stage diffusion sub-regions and the N first initial diffusion sub-regions is calculated respectively, and then it is determined whether the difference between the mean value of the pixel gradients in the N first-stage diffusion sub-regions and the mean value of the pixel gradients in the N first initial diffusion sub-regions meets the preset difference value. If so, it indicates that the edge of the clear region and the blurred region has been reached, and the preset diffusion stop condition is met, and the N first-stage diffusion sub-regions are used as the N first effective sub-regions.

[0056] If not, it indicates that the edge of the clear area and the blurred area has not been reached. At this time, the edge diffusion of the N first-stage diffusion sub-areas is performed again according to the preset area diffusion step length until the preset diffusion stop condition is met to obtain the N first effective sub-areas.

[0057] The above steps calculate the pixel gradient, diffuse and refine step by step, and finally determine the effective area set of each image. This method ensures that only high-quality areas in the image are identified and extracted, avoiding the influence of fuzzy areas on subsequent target extraction and spatial analysis.

[0058] Furthermore, step S200 of the embodiment of the present application further includes:

[0059] Performing deep learning processing on the M sets of valid image regions to extract M clusters of objects of interest, each cluster of objects of interest including K sets of objects of interest;

[0060] Traversing the M clusters of interest to perform spatial coordinate mapping analysis to determine M target spatial coordinate clusters;

[0061] Spatial edge analysis is performed based on the M target space coordinate clusters and the M device layout positions to obtain the M effective image acquisition space sets.

[0062] In one embodiment, the M sets of valid image regions obtained in the previous step are used as input to the deep learning model. The valid region set contains high-quality parts selected from images collected from different devices. Pre-trained target detection models (such as YOLO) are used to detect the objects of interest in the image. These models should be able to detect and classify different targets in complex jungle environments, such as trees, bushes, stones and other jungle features.

[0063] Preferably, target detection is performed on each effective area of ​​each image to extract a cluster of targets of interest. Each cluster of targets of interest includes K sets of targets of interest (K is an integer greater than or equal to 1), which represents a set of targets identified from each viewing angle in each monitoring scene. Targets of interest include trees, bushes, grass, and stones, etc. The position of each target of interest in the target jungle is determined based on the M clusters of targets of interest, and the M target space coordinate clusters are obtained.

[0064] Preferably, the two-dimensional pixel coordinates in the image are mapped to coordinates in the actual three-dimensional space according to the position, layout angle and imaging geometric parameters of each image acquisition device. This step maps the target from the image plane to the actual jungle environment space through reverse projection or multi-view geometry method. Preferably, the position of the target of interest in the target jungle can also be manually determined to obtain the corresponding spatial coordinates.

[0065] Optionally, taking the M device layout positions as M space centers and combining the M target space coordinate clusters, extract the target space coordinate with the farthest distance from each angle to the M space centers and use it as an edge point of the space edge, and then connect the multiple edge points obtained to obtain M valid image acquisition space sets. Preferably, the edge of the target cluster is analyzed in combination with the layout position of each device to ensure that the coverage of the device can fully capture the boundary of the target cluster. If the boundaries of some target clusters are located in the acquisition blind spot of the device, the system will prompt to readjust the device layout angle or add new devices. Through the above steps, it is ensured that all key targets in the complex jungle environment are accurately detected and mapped, providing reliable data support for subsequent three-dimensional space reconstruction.

[0066] Step S300: using the K monitoring scenes as indexes, performing overlapping space identification on the M valid image acquisition space sets and the M equipment deployment positions, and determining K image acquisition overlapping space sets, wherein each image acquisition overlapping space includes multiple overlapping objects of interest and multiple superimposed pixel gradients;

[0067] In one embodiment, the K monitoring scenes are used as indexes, and the M effective image acquisition space sets are combined to aggregate the effective image acquisition spaces belonging to the same monitoring scene into one set to obtain K effective image acquisition space sets. Thus, the consistency and synchronization of information such as device position and angle in each monitoring scene is achieved.

[0068] The first effective image acquisition space set is extracted from the K effective image acquisition space sets. Since the M device layout positions are the spatial centers of the corresponding spaces in the first effective image acquisition space set, the first effective image acquisition space set can be placed in the same coordinate system according to the M device layout positions. Overlapping space identification is performed on the first effective image acquisition space set to obtain the K image acquisition overlapping space sets.

[0069] Preferably, in combination with the M target space coordinate clusters corresponding to the M clusters of interest in the above step S200, and the K sets of image acquisition overlapping spaces, the targets of interest existing in each image acquisition overlapping space can be determined and used as multiple overlapping targets of interest in each image acquisition overlapping space. Since each overlapping target of interest may exist in multiple valid image acquisition spaces, and the pixel gradients in different valid image acquisition spaces are different, the larger the value after the multiple pixel gradients in the multiple valid image acquisition spaces are superimposed, the higher the clarity of the overlapping target of interest in the image acquisition overlapping space, and the higher the accuracy and tolerance of spatial stitching using this as a stitching path. Therefore, by obtaining each image acquisition overlapping space including multiple overlapping targets of interest and multiple superimposed pixel gradients, the technical effect of paving the way for the subsequent determination of the spatial stitching path is achieved.

[0070] Step S400: traversing the K image acquisition overlapping space sets to perform spatial fusion recognition of the target of interest, determining a spatial splicing path according to the recognition result, and obtaining a set of K overlapping spatial splicing paths;

[0071] Further, the K image acquisition overlapping space sets are traversed to perform spatial fusion recognition of the target of interest, and a spatial stitching path is determined according to the recognition result to obtain a set of K overlapping spatial stitching paths. In the embodiment of the present application, step S400 further includes:

[0072] Randomly extracting a first image acquisition overlapping space from the K image acquisition overlapping space sets, wherein the first image acquisition overlapping space includes a plurality of overlapping objects of interest and a plurality of superimposed pixel point gradients;

[0073] Extracting an overlapping object of interest located at the edge of the space and corresponding to the minimum value of the gradient of the superimposed pixel point from the multiple overlapping objects of interest, and taking it as the starting point of the first path;

[0074] Determine a first starting neighborhood of the first path starting point based on the distances of the multiple overlapping interested objects to the first path starting point, wherein the first starting neighborhood is a set of overlapping interested objects whose distances to the first path starting point meet a preset distance threshold;

[0075] Taking the overlapping interested object corresponding to the maximum value of the gradient of the superimposed pixel point in the first starting neighborhood as the first path stage point, and constructing the neighborhood of the first path stage point in combination with the multiple overlapping interested objects, and determining the next first path stage point according to the construction result, until reaching the edge of the first image acquisition overlapping space, and obtaining multiple first path stage points;

[0076] Connecting the first path starting point and the plurality of first path stage points in sequence according to the order in which they are obtained, to obtain a first overlapping space splicing path;

[0077] Perform spatial stitching path identification on the K image acquisition overlapping space sets to obtain K overlapping space stitching path sets.

[0078] In one embodiment, according to the K overlapping image acquisition spaces, each overlapping image acquisition space includes multiple overlapping objects of interest and multiple superimposed pixel gradients, after the two-dimensional images acquired by different image acquisition devices in each overlapping image acquisition space are converted into three-dimensional point clouds, the paths for splicing the three-dimensional point clouds formed after the image conversion by different image acquisition devices in each overlapping image acquisition space are determined, and the K overlapping space splicing path sets are obtained. The technical effect of facilitating the effective fusion of data acquired by multiple devices and perspectives in a jungle environment and providing complete spatial path planning for subsequent three-dimensional reconstruction is achieved.

[0079] Preferably, one of the K overlapping space sets is randomly selected as the first image acquisition overlapping space for initial planning of the spatial stitching path. The first image acquisition overlapping space contains multiple overlapping objects of interest and multiple superimposed pixel gradients. This information will help determine the target distribution and image quality in the overlapping area.

[0080] In the overlapping space of the first image acquisition, the objects of interest located at the edge of the space are preferentially extracted. These objects usually represent geographical boundaries, key feature points in the jungle, or objects in the boundary area of ​​the image. Then, the object with the smallest superimposed pixel gradient is selected as the starting point of the first path. The object with the smallest pixel gradient represents that the image quality of the area is low or there are fewer details, which is suitable as the starting point of the path.

[0081] Taking the starting point of the first path as the center, calculate the distance between the starting point and other overlapping interested targets. According to the preset distance threshold (the maximum distance from the overlapping interested targets in the neighborhood to the starting center of the first path pre-set by those skilled in the art), determine the target that is closer to the starting point as the first starting neighborhood. Preferably, the targets in the first starting neighborhood are a set of interested targets that meet the distance threshold requirements, and these targets will serve as potential path stage points.

[0082] Then, in the first starting neighborhood, the target corresponding to the maximum value of the superimposed pixel gradient is selected as the first path stage point. A higher gradient value means better image quality and clearer edges, so it is suitable as a path midpoint. Based on the same principle, the neighborhood construction process is repeated for the selected path stage points, and the path neighborhood is continued to be expanded according to the distance from other overlapping targets of interest. The distance and gradient difference between each stage point are used to determine the selection of the next path stage point. The next path stage point is continuously selected until the edge of the first image acquisition overlapping space is reached, and multiple first path stage points are obtained. In the order obtained, multiple first path stage points are connected in sequence to generate a first overlapping space splicing path. This path represents a complete spatial splicing process from the edge to the edge of the overlapping space. Repeat the above path planning process for each overlapping space set, determine the starting point, path stage point and path extension of each overlapping space, until the splicing path of all overlapping spaces is obtained, that is, the K overlapping space splicing path sets. Each splicing path set represents a part of the jungle environment, and the splicing is completed by images taken by different devices and different angles. The technical effect of improving the accuracy of three-dimensional space reconstruction is achieved.

[0083] Step S500: collecting node monitoring scene features of the jungle environment monitoring network in the current reconstruction window, matching them with the K monitoring scenes to obtain a matching monitoring scene, and determining a matching overlapping space splicing path set according to the matching monitoring scene and the K overlapping space splicing path sets;

[0084] In a possible embodiment, the current reconstruction window is the time point for three-dimensional reconstruction of the target jungle, and the environmental conditions of the jungle environment monitoring network are collected at this time point to obtain node monitoring scene features. The node monitoring scene features are used to describe the environmental conditions of the target jungle in the current reconstruction window, including factors such as lighting, weather, terrain, and vegetation. Optionally, the feature data of these nodes are obtained using an image acquisition device or other sensors that have been set up in advance.

[0085] Analyze the monitoring scene features collected in the current reconstruction window and compare them with the feature sets of the K monitoring scenes defined in the previous step. Use the cosine similarity calculation formula to determine the similarity between the node monitoring scene features and the K jungle environment feature sets of the K monitoring scenes, and use the monitoring scene corresponding to the maximum similarity as the matching monitoring scene. According to the mapping relationship between the monitoring scene and the K overlapping space stitching path sets, determine the corresponding matching overlapping space stitching path set based on the matching monitoring scene search. A technical effect of providing a stitching path for subsequent three-dimensional space reconstruction is achieved.

[0086] Step S600: extracting the captured images of M image acquisition devices in the current reconstruction window to obtain M real-time captured images, performing three-dimensional reconstruction based on the M real-time captured images, and splicing the reconstructed space in combination with the matching overlapping space splicing path set to determine the three-dimensional space of the target reconstruction environment of the target jungle.

[0087] Furthermore, step S600 in the embodiment of the present application further includes:

[0088] Obtaining M window angle information of M image acquisition devices under the current reconstruction window;

[0089] Based on the M device deployment angles and the M angle tolerance intervals, the M window angle information is authenticated to obtain M authentication results;

[0090] If the M authentication results are all authenticated, obtaining an image acquisition instruction;

[0091] The M image acquisition devices are used to acquire images based on the image acquisition instruction to obtain the M real-time acquired images.

[0092] In a possible embodiment, images collected by M image acquisition devices in the current reconstruction window are extracted to obtain the M real-time collected images. The M real-time collected images reflect the real-time environment of the target jungle. Three-dimensional reconstruction is performed based on the M real-time collected images, and the overlapping spaces in the reconstruction space are spliced ​​in combination with the matching overlapping space splicing path set to determine the target reconstruction environment three-dimensional space of the target jungle. The technical effect of improving the reliability and accuracy of three-dimensional space reconstruction is achieved.

[0093] Preferably, the M image acquisition devices are connected, the camera angle information of the M image acquisition devices is extracted, and the M window angle information of the M image acquisition devices under the current reconstruction window is obtained. Then, based on the M device layout angles and M angle tolerance intervals (the angle range within which the cameras of the M image acquisition devices can be offset set by those skilled in the art), the M window angle information is authenticated to obtain M authentication results.

[0094] Optionally, when the difference between the M window angle information and the M device layout angles satisfies the M angle tolerance intervals, the authentication result is passed authentication. When the difference between the M window angle information and the M device layout angles does not satisfy the M angle tolerance intervals, the authentication result is failed authentication. If the M authentication results are all passed authentication, an image acquisition instruction is obtained, and image acquisition of the M image acquisition devices is performed based on the image acquisition instruction to obtain the M real-time acquired images. Thus, the technical effect of improving the reliability of the acquired images is achieved.

[0095] In one embodiment, the M real-time collected images are pre-processed, including denoising, contrast enhancement and image correction, to ensure that the quality of each image meets the requirements of three-dimensional reconstruction. Feature points are extracted for each real-time image. The SIFT algorithm can be used to extract key points in the image, and these feature points will be used for matching and alignment between images. Feature point matching is performed between the M real-time images, and the feature points under different perspectives are matched by the FLANN algorithm. The successfully matched feature points will be used to generate a three-dimensional space point cloud. Based on the matched feature points, a multi-view stereo reconstruction (Multi-View Stereo, MVS) technology is used to generate a three-dimensional point cloud. This step combines the different perspective information of the M real-time collected images to generate a three-dimensional geometric structure of the jungle environment. Exemplarily, a sparse point cloud model is first generated, which reflects the three-dimensional spatial position of the main features in the jungle and forms a preliminary geometric framework of the entire environment. Based on the sparse point cloud, a multi-view stereo algorithm is used to reconstruct a dense point cloud to generate a more refined three-dimensional structure. The dense point cloud will contain more details, reflecting the tiny features of elements such as plants and terrain in the jungle.

[0096] Furthermore, the matching overlapping space stitching path set obtained from step S500 includes the stitching path information of multiple overlapping spaces of different image acquisition devices. The generated three-dimensional point cloud is stitched using the matching overlapping space stitching path set to seamlessly connect the point cloud data in the overlapping areas acquired by M image acquisition devices. Ensure that the point cloud models generated by each device are aligned in space to form a continuous three-dimensional reconstruction model. Process the redundant points in the overlapping areas to eliminate duplicate data caused by shooting with multiple devices. Optimize the stitching quality of these overlapping areas through weighted averaging or feature fusion technology to ensure the continuity and smoothness of the visual effect. Thus, a complete three-dimensional spatial model of the jungle is constructed. This model includes the terrain,

[0097] Step S600 generates a target 3D space model of the jungle environment through real-time image acquisition, 3D reconstruction and spatial splicing. This process ensures that the generated 3D model has high accuracy and complete details through the integration of image preprocessing, feature point matching, point cloud generation and splicing paths, and finally presents a complete 3D space of the jungle environment.

[0098] In summary, the method for reconstructing a complex jungle environment three-dimensional space based on image recognition provided by this application has the following technical effects:

[0099] 1. This application achieves high-precision reconstruction of complex jungle environments through efficient multi-device image acquisition, real-time 3D reconstruction, automated spatial stitching, and flexible scene adaptation. This method not only improves the accuracy and efficiency of the 3D model, but also enhances the adaptability to complex environments.

[0100] 2. By identifying and fusing the targets of interest in the overlapping space, and automatically generating the stitching path based on pixel gradient calculation and spatial feature matching, it effectively reduces manual intervention, improves the efficiency of spatial stitching, and achieves the technical effect of improving the accuracy of three-dimensional space reconstruction.

[0101] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0102] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A method for reconstructing three-dimensional space in a complex jungle environment based on image recognition, characterized in that: The method comprises: Deploy an environmental monitoring network based on the initial detection information of the target jungle to obtain a jungle environmental monitoring network, wherein the jungle environmental monitoring network includes M image acquisition devices, M device deployment positions, and M device deployment angles, where M is an integer greater than or equal to 1; Interacting the basic device information of the M image acquisition devices, determining M effective image acquisition space sets of the M image acquisition devices under K monitoring scenes and the M device deployment angles, each effective image acquisition space is an area range of an image acquisition device in a monitoring scene where the image quality meets the requirements, wherein K is an integer greater than or equal to 1; Using the K monitoring scenes as indexes, performing overlapping space identification on the M valid image acquisition space sets and the M equipment deployment positions, and determining K image acquisition overlapping space sets, wherein each image acquisition overlapping space includes multiple overlapping objects of interest and multiple superimposed pixel point gradients; Traversing the K image acquisition overlapping space sets to perform spatial fusion recognition of the target of interest, determining a spatial splicing path according to the recognition result, and obtaining a set of K overlapping spatial splicing paths; Collecting node monitoring scene features of the jungle environment monitoring network in the current reconstruction window, matching them with the K monitoring scenes to obtain a matching monitoring scene, and determining a matching overlapping space splicing path set according to the matching monitoring scene and the K overlapping space splicing path sets; The captured images of M image acquisition devices in the current reconstruction window are extracted to obtain M real-time captured images, three-dimensional reconstruction is performed based on the M real-time captured images, and the reconstructed space is spliced ​​in combination with the matching overlapping space splicing path set to determine the three-dimensional space of the target reconstruction environment of the target jungle.

2. The method for reconstructing a complex jungle environment three-dimensional space based on image recognition as claimed in claim 1, characterized in that: Interacting the basic device information of the M image acquisition devices to determine M effective image acquisition space sets of the M image acquisition devices under K monitoring scenes and the M device deployment angles, including: Obtaining K jungle environment feature sets of the K monitoring scenes; Interacting with the jungle environment monitoring network to obtain M sets of basic information of the M image acquisition devices, and performing test image acquisition in combination with the M device layout angles and the K jungle environment feature sets to obtain M test image sets, wherein each test image set includes K test images, and each test image corresponds to a monitoring scene; Traversing the M test image sets to perform pixel point gradient calculations, and determining M image valid area sets according to the calculation results; The objects of interest are extracted from the M sets of effective image regions, and spatial coordinate mapping analysis is performed in combination with the extraction results, and the M sets of effective image acquisition spaces are determined according to the analysis results.

3. The method for reconstructing a complex jungle environment three-dimensional space based on image recognition as claimed in claim 2, characterized in that: Traversing the M test image sets to perform pixel point gradient calculations, and determining M image valid area sets according to the calculation results, including: Randomly extracting a first test image from the set of M test images, and performing gradient calculation on pixel points in the first test image to obtain a plurality of first pixel point gradients; Taking the pixel points corresponding to the N first pixel point gradients located in the first N positions among the multiple first pixel point gradients as the center points of N first effective sub-regions, taking the center points of the N first effective sub-regions as the starting point, performing effective sub-region diffusion according to a preset area diffusion step length until a preset diffusion stop condition is met, and obtaining N first effective sub-regions, wherein N is an integer greater than or equal to 1, and the preset area diffusion step length is the maximum pixel point gradient difference allowed in a single diffusion; Taking a union of the N first valid sub-regions to obtain a first image valid region; Pixel point gradient calculation is performed on the M test image sets according to the preset area diffusion step length to determine the M image valid area sets.

4. The method for reconstructing a three-dimensional space of a complex jungle environment based on image recognition as claimed in claim 3, characterized in that: include: Taking the center points of the N first effective sub-regions as starting points, effective sub-region diffusion is performed according to a preset region diffusion step length to obtain N first initial diffusion sub-regions; Perform edge diffusion on the N first initial diffusion sub-regions again according to the preset regional diffusion step length to obtain N first-stage diffusion sub-regions; Determine whether a difference between a pixel point gradient mean value of the N first stage diffusion sub-regions and a pixel point gradient mean value of the N first initial diffusion sub-regions satisfies a preset difference value, and if so, the preset diffusion stop condition is satisfied, and the N first stage diffusion sub-regions are used as N first effective sub-regions; If not, edge diffusion is performed again on the N first-stage diffusion sub-regions according to the preset region diffusion step length until a preset diffusion stop condition is met, thereby obtaining the N first effective sub-regions.

5. The method for reconstructing a complex jungle environment three-dimensional space based on image recognition as claimed in claim 2, characterized in that: include: Performing deep learning processing on the M sets of valid image regions to extract M clusters of objects of interest, each cluster of objects of interest including K sets of objects of interest; Traversing the M clusters of interest to perform spatial coordinate mapping analysis to determine M target spatial coordinate clusters; Spatial edge analysis is performed based on the M target space coordinate clusters and the M device layout positions to obtain the M effective image acquisition space sets.

6. The method for reconstructing a three-dimensional space of a complex jungle environment based on image recognition as claimed in claim 1, characterized in that: Traversing the K image acquisition overlapping space sets to perform spatial fusion recognition of the target of interest, determining the spatial splicing path according to the recognition result, and obtaining K overlapping spatial splicing path sets, including: Randomly extracting a first image acquisition overlapping space from the K image acquisition overlapping space sets, wherein the first image acquisition overlapping space includes a plurality of overlapping objects of interest and a plurality of superimposed pixel point gradients; Extracting an overlapping object of interest located at the edge of the space and corresponding to the minimum value of the gradient of the superimposed pixel point from the multiple overlapping objects of interest, and taking it as the starting point of the first path; Determine a first starting neighborhood of the first path starting point based on the distances of the multiple overlapping interested objects to the first path starting point, wherein the first starting neighborhood is a set of overlapping interested objects whose distances to the first path starting point meet a preset distance threshold; Taking the overlapping interested object corresponding to the maximum value of the gradient of the superimposed pixel point in the first starting neighborhood as the first path stage point, and constructing the neighborhood of the first path stage point in combination with the multiple overlapping interested objects, and determining the next first path stage point according to the construction result, until reaching the edge of the first image acquisition overlapping space, and obtaining multiple first path stage points; Connecting the first path starting point and the plurality of first path stage points in sequence according to the order in which they are obtained, to obtain a first overlapping space splicing path; Perform spatial stitching path identification on the K image acquisition overlapping space sets to obtain K overlapping space stitching path sets.

7. The method for reconstructing a complex jungle environment three-dimensional space based on image recognition as claimed in claim 2, characterized in that: Each of the K jungle environment feature sets includes lighting conditions, weather conditions and monitoring seasons.

8. The method for reconstructing a complex jungle environment three-dimensional space based on image recognition as claimed in claim 1, characterized in that: include: Obtaining M window angle information of M image acquisition devices under the current reconstruction window; Based on the M device deployment angles and the M angle tolerance intervals, the M window angle information is authenticated to obtain M authentication results; If the M authentication results are all authenticated, obtaining an image acquisition instruction; The M image acquisition devices are used to acquire images based on the image acquisition instruction to obtain the M real-time acquired images.

Images