Method and system for monitoring populations of migratory birds

Abstract

This invention proposes a method and system for identifying groups of migratory birds in their habitats. The method includes: segmenting a migratory bird monitoring image into a predetermined number of smaller images; inputting the smaller images into a trained migratory bird target recognition model to initially label each target region in the smaller images based on the target recognition results; determining whether duplicate target regions exist within overlapping regions based on the contour coordinates of overlapping regions; if duplicate target regions exist within overlapping regions, deleting one of the duplicate target regions contained in any two adjacent smaller images; inputting the image to be identified into a trained fine-grained migratory bird recognition model to obtain the species corresponding to each migratory bird target included in each target region; performing secondary labeling on each migratory bird target included in each target region, and summarizing the secondary labeling results from all images to be identified. This invention can accurately identify the species of each migratory bird target contained in each target region, and thus accurately calculate the total number of each migratory bird species in the target habitat.

Description

Technical Field

[0001] This invention relates to the field of biometrics, and in particular to a method and system for identifying groups of migratory birds in their habitats. Background Technology

[0002] Currently, with the progress and development of human society, people's awareness of protecting the natural ecological environment is constantly increasing. Birds, as friends of humans, have been listed as protected objects, so the study of birds is of great significance.

[0003] Bird migration is an adaptive behavior to the periodic changes in environmental factors, and it is of great significance to population reproduction, evolution, and the maintenance of biodiversity. Traditional methods mainly fall into two categories: one is through manual observation, but this method of identifying and monitoring rare migratory birds suffers from low efficiency and high costs. The other method combines deep learning technology, which can identify the species and numbers of migratory birds. However, the accuracy of identification requires a large number of migratory bird samples. Using platform video monitoring to capture the most real-time and accurate migratory bird samples consumes a significant amount of manpower, and is still affected by the timing of bird migration and the activities of the birds, so the identification accuracy has not yet reached the ideal level. Summary of the Invention

[0004] Based on this, the purpose of this invention is to propose a method and system for group identification of migratory bird habitats, aiming to solve the problem of low accuracy in traditional methods of monitoring and identifying migratory birds.

[0005] A method for group identification in monitoring migratory bird habitats, proposed according to the present invention, is applied to a migratory bird monitoring and management platform. The method includes:

[0006] Every first preset time interval, migratory bird monitoring images of the target habitat are acquired, and the migratory bird monitoring images are divided into a preset number of small images, with any two adjacent small images having an overlapping portion;

[0007] The small image is input into a trained migratory bird target recognition model to perform initial labeling on each target region in the small image based on the target recognition results, and each target region includes at least one migratory bird target;

[0008] Obtain the contour coordinates of the overlapping region between any two adjacent small images, and determine whether there is a duplicate target region within the overlapping region based on the contour coordinates of the overlapping region;

[0009] If there are duplicate target regions within the overlapping area, then delete one of the duplicate target regions contained in any two adjacent small images to obtain the image to be identified after deleting the duplicate target regions.

[0010] The image to be identified, corresponding to each of the small images, is input into the trained fine-grained migratory bird recognition model, so that the fine-grained migratory bird recognition model can identify each first-marked target region in the image to be identified, and obtain the types of each migratory bird target included in each target region.

[0011] Each migratory bird target in each target area is secondary-labeled according to the species corresponding to each migratory bird target included in each target area, and the secondary labeling results in all the images to be identified are summarized to obtain the total number corresponding to each migratory bird species.

[0012] In summary, based on the aforementioned method for group identification of migratory bird habitats, the method first locks the target area using a large-scale migratory bird monitoring image database, and then performs targeted target species identification for each target area. This accurately identifies the species of each migratory bird target contained in each target area, and then precisely calculates the total number of each migratory bird species in the target habitat. Specifically, the method first monitors migratory bird monitoring images of the target habitat in real time and divides these images into multiple smaller images, ensuring that adjacent smaller images overlap. Then, the smaller images are used for target area identification, and the identified target areas are initially marked. Next, it is determined whether any overlapping areas contain duplicate target areas. If so, only one target area is retained. The image to be identified after removing duplicate target areas is then input into a fine-grained migratory bird identification model, thereby identifying the species of each migratory bird target contained in each target area, and thus accurately obtaining the total number of each migratory bird species.

[0013] Further, the step of acquiring migratory bird monitoring images of the target habitat at first preset time intervals, and dividing the migratory bird monitoring images into a preset number of smaller images, wherein any two adjacent smaller images have overlapping portions, includes:

[0014] The size information of the migratory bird monitoring image is obtained, and the position coordinates of all first cutting lines are obtained according to the size information and the first preset spacing value. The position coordinates of all second cutting lines are obtained according to the position coordinates of the first cutting lines and the second preset value. The first cutting lines and the second cutting lines constitute multiple overlapping areas.

[0015] The migratory bird monitoring image is segmented into multiple smaller images based on the position coordinates of all the first cutting lines and all the position coordinates of all the second cutting lines.

[0016] Further, the step of inputting the small image into a trained migratory bird target recognition model, and initially labeling each target region in the small image according to the target recognition result, wherein each target region includes at least one migratory bird target, includes:

[0017] Acquire multiple historical migratory bird images, and obtain the location information of all known target areas contained in each historical migratory bird image based on the historical migratory bird images;

[0018] Each historical migratory bird image is labeled based on the location information of all known target areas contained in each historical migratory bird image, and the migratory bird target recognition model is trained based on the labeled historical migratory bird images.

[0019] Further, the step of obtaining the contour coordinates of the overlapping region between any two adjacent small images, and determining whether there is a duplicate target region within the overlapping region based on the contour coordinates of the overlapping region, includes:

[0020] Obtain the contour coordinates of each target region marked for the first time, and traverse all target regions based on the contour coordinates of the overlapping region and the contour coordinates of each target region to determine whether there is a target region within the overlapping region.

[0021] Further, the step of inputting the image to be identified corresponding to each of the small images into the trained fine-grained migratory bird recognition model, so that the fine-grained migratory bird recognition model can identify each initially labeled target region in the image to be identified and obtain the species corresponding to each migratory bird target included in each target region, includes:

[0022] Acquire images of migratory birds of known types, where each migratory bird target in the image corresponds to a type of migratory bird. Extract features from each migratory bird target in the image to obtain head feature information, neck feature information, torso feature information, and leg feature information corresponding to each migratory bird target.

[0023] The fine-grained migratory bird recognition model is trained based on the head feature information, neck feature information, trunk feature information, leg feature information, and migratory bird type corresponding to each migratory bird target.

[0024] The trained fine-grained migratory bird recognition model identifies each target region in the image to be recognized based on the head feature information, the neck feature information, the trunk feature information, and the leg feature information.

[0025] In another aspect, the present invention provides a group identification system for monitoring migratory bird habitats, the system comprising:

[0026] The image segmentation module is used to acquire migratory bird monitoring images of the target habitat at a first preset time interval, and to segment the migratory bird monitoring images into a preset number of small images, with any two adjacent small images having an overlapping portion.

[0027] The initial labeling module is used to input the small image into the trained migratory bird target recognition model, so as to perform initial labeling on each target region in the small image according to the target recognition result, and each target region includes at least one migratory bird target;

[0028] The duplicate region detection module is used to obtain the contour coordinates of the overlapping region between any two adjacent small images, and to determine whether there is a duplicate target region in the overlapping region based on the contour coordinates of the overlapping region.

[0029] The duplicate region deletion module is used to delete one of the duplicate target regions contained in any two adjacent small images if there are duplicate target regions in the overlapping region, so as to obtain the image to be identified after deleting the duplicate target regions.

[0030] The migratory bird species identification module is used to input the image to be identified corresponding to each of the small images into the trained migratory bird fine-grained identification model, so that the migratory bird fine-grained identification model can identify each first-marked target region in the image to be identified, and obtain the species corresponding to each migratory bird target included in each target region;

[0031] The secondary labeling module is used to perform secondary labeling on each migratory bird target included in each target area according to the species corresponding to each migratory bird target included in each target area, and to summarize the secondary labeling results in all the images to be identified to obtain the total number corresponding to each migratory bird species.

[0032] Furthermore, the image segmentation module also includes:

[0033] The cutting line acquisition unit is used to acquire the size information of the migratory bird monitoring image, and acquire the position coordinates of all first cutting lines according to the size information and a first preset spacing value, and acquire the position coordinates of all second cutting lines according to the position coordinates of the first cutting lines and a second preset value, wherein the first cutting lines and the second cutting lines constitute multiple overlapping areas;

[0034] The cutting execution unit is used to segment the migratory bird monitoring image into multiple smaller images based on the position coordinates of all the first cutting lines and all the position coordinates of the second cutting lines.

[0035] Furthermore, the initial marking module also includes:

[0036] The historical migratory bird image acquisition unit is used to acquire multiple historical migratory bird images and, based on the historical migratory bird images, acquire the location information of all known target areas contained in each historical migratory bird image;

[0037] The historical migratory bird image annotation unit is used to annotate each historical migratory bird image according to the location information of all known target areas contained in each historical migratory bird image, and to train the migratory bird target recognition model based on the annotated historical migratory bird images.

[0038] Furthermore, the repeating region detection module also includes:

[0039] The contour coordinate acquisition unit is used to acquire the contour coordinates of each target region initially marked, and to traverse all target regions based on the contour coordinates of the overlapping regions and the contour coordinates of each target region to determine whether a target region exists within the overlapping regions.

[0040] Furthermore, the migratory bird species identification module also includes:

[0041] The feature extraction unit is used to acquire images of migratory birds of known types, wherein each migratory bird target in the migratory bird image corresponds to a type of migratory bird, and to extract features from each migratory bird target in the migratory bird image to obtain head feature information, neck feature information, torso feature information and leg feature information corresponding to each migratory bird target;

[0042] The fine-grained recognition model training unit is used to train the migratory bird fine-grained recognition model based on the head feature information, neck feature information, trunk feature information, leg feature information corresponding to each migratory bird target, and the migratory bird type corresponding to each migratory bird target;

[0043] The migratory bird species identification execution unit is used by the trained fine-grained migratory bird identification model to identify each target region in the image to be identified based on the head feature information, the neck feature information, the trunk feature information, and the leg feature information.

[0044] In another aspect, the present invention provides a storage medium, including storage of one or more programs that, when executed, implement the above-described method for group identification of migratory bird habitats.

[0045] In another aspect, the present invention provides a computer device, the computer device including a memory and a processor, wherein:

[0046] The memory is used to store computer programs;

[0047] When the processor executes the computer program stored in the memory, it implements the above-described method for group identification of migratory bird habitats.

[0048] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by means of embodiments of the invention. Attached Figure Description

[0049] Figure 1 This is a flowchart of the method for group identification of migratory bird habitats proposed in the first embodiment of the present invention;

[0050] Figure 2 This is a schematic diagram of migratory bird target recognition in the first embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram of the migratory bird species identification results in the first embodiment of the present invention;

[0052] Figure 4 This is a schematic diagram of the structure of the group identification system for monitoring migratory bird habitats proposed in the second embodiment of the present invention.

[0053] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0054] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0055] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0056] Please see Figure 1 The diagram shows a flowchart of a method for monitoring and identifying migratory bird habitats according to a first embodiment of the present invention. The method includes steps S01 to S06, wherein:

[0057] Step S01: Acquire migratory bird monitoring images of the target habitat at a first preset time interval, and divide the migratory bird monitoring images into a preset number of small images, with any two adjacent small images having an overlapping portion;

[0058] It should be noted that the target habitats are generally places where migratory birds can easily inhabit, such as lakesides and wetlands. The migratory bird monitoring images are captured by cameras installed near the habitats. Since the target habitats generally occupy a large area, the captured migratory bird monitoring images are usually images with very large memory. In order to improve the recognition efficiency and accuracy, the migratory bird monitoring images are first divided into multiple smaller images. At the same time, in order to avoid data loss and duplicate counting, it is necessary to ensure that each small image has an overlapping part during the segmentation process.

[0059] In some embodiments of the present invention, in order to segment migratory bird monitoring images, the size information of the migratory bird monitoring images is first obtained, and the position coordinates of all first cutting lines are obtained according to the size information and a first preset spacing value. The position coordinates of all second cutting lines are obtained according to the position coordinates of the first cutting lines and a second preset value. The first cutting lines and the second cutting lines constitute multiple overlapping areas. It should be noted that there are multiple first cutting lines and second cutting lines in both the horizontal and vertical directions, and the area between adjacent first cutting lines and second cutting lines is the overlapping part.

[0060] Then, the migratory bird monitoring image is segmented into multiple smaller images based on the position coordinates of all the first cutting lines and all the position coordinates of all the second cutting lines. If there are no overlapping parts during segmentation, data loss can easily occur, affecting the accuracy of species identification. At the same time, if the same migratory bird target is segmented into two parts, it is not easy to eliminate duplicate migratory bird targets.

[0061] To monitor the target habitat continuously in real time, the first preset time is generally set to one day or one week, meaning the target habitat is monitored once a day or once a week.

[0062] Step S02: Input the small image into the trained migratory bird target recognition model, so as to perform initial labeling on each target region in the small image according to the target recognition result, and each target region includes at least one migratory bird target;

[0063] In this step, in order to build a trained migratory bird target recognition model, it is first necessary to acquire multiple historical migratory bird images, and then obtain the location information of all known target areas contained in each historical migratory bird image based on the location information of all known target areas contained in each historical migratory bird image. Then, each historical migratory bird image is labeled based on the location information of all known target areas contained in each historical migratory bird image, and the migratory bird target recognition model is trained based on the labeled historical migratory bird images.

[0064] Please see Figure 2 The image shown is a schematic diagram for identifying migratory birds. Figure 2It can be seen that the trained migratory bird target recognition model can identify all regions in the image where migratory birds exist, and then use square boxes to mark the identified regions for the first time.

[0065] Furthermore, it should be noted that in the actual identification process, multiple closely spaced migratory birds may be identified as a single migratory bird target due to shooting angle issues, which would affect the accuracy of species counting. Based on this, this embodiment first uses a migratory bird target identification model to accurately determine the target area where each migratory bird target exists, and then performs accurate species identification on each target area where a migratory bird target exists, thereby greatly improving the identification accuracy.

[0066] Step S03: Obtain the contour coordinates of the overlapping region between any two adjacent small images, and determine whether there is a duplicate target region in the overlapping region based on the contour coordinates of the overlapping region;

[0067] It should be noted that, in order to avoid overlapping parts causing duplicate counting, it is necessary to accurately obtain the contour coordinates of each target region marked for the first time, that is, the contour coordinates of each marker box marked for the first time. Based on the contour coordinates of the overlapping region and the contour coordinates of each target region, all target regions are traversed to determine whether there is a target region in the overlapping region. The contour coordinates of the overlapping region are obtained from the position coordinates of the first cutting line and the second cutting line that constitute the overlapping region.

[0068] Step S04: If there are duplicate target regions in the overlapping area, delete one of the duplicate target regions contained in any two adjacent small images to obtain the image to be identified after deleting the duplicate target regions.

[0069] Understandably, when overlapping target regions are identified, only one of them is retained to prevent repeated identification during subsequent category identification.

[0070] Step S05: Input the image to be identified corresponding to each of the small images into the trained fine-grained migratory bird recognition model, so that the fine-grained migratory bird recognition model can identify each first-marked target region in the image to be identified, and obtain the types of each migratory bird target included in each target region;

[0071] It should be noted that, in order to accurately identify the species of each migratory bird in the image to be identified and obtain images of migratory birds of known species, each migratory bird in the image corresponds to a type of migratory bird, and feature extraction is performed on each migratory bird in the image to obtain head feature information, neck feature information, torso feature information and leg feature information corresponding to each migratory bird.

[0072] The fine-grained migratory bird identification model is trained based on the head, neck, trunk, and leg features corresponding to each migratory bird target, as well as the migratory bird type corresponding to each migratory bird target. By incorporating the head, neck, trunk, and leg features of migratory birds into species identification, the accuracy of migratory bird species identification can be greatly improved.

[0073] Finally, the trained fine-grained migratory bird recognition model is used to identify each target region in the image to be recognized based on the head feature information, the neck feature information, the trunk feature information, and the leg feature information.

[0074] Step S06: Based on the species of each migratory bird included in each target area, perform secondary labeling on each migratory bird included in each target area, and summarize the secondary labeling results in all images to be identified to obtain the total number corresponding to each migratory bird species.

[0075] It should be noted that in this step, please refer to [link / reference]. Figure 3 The diagram shows the results of migratory bird species identification. Since the secondary labeling object is each migratory bird target and the labeling result is the migratory bird species identified for each migratory bird target, we only need to summarize all the secondary labeling boxes and the migratory bird species corresponding to the secondary labeling boxes to obtain the total number of each migratory bird species.

[0076] In summary, based on the aforementioned method for group identification of migratory bird habitats, the method first locks the target area using a large-scale migratory bird monitoring image database, and then performs targeted target species identification for each target area. This accurately identifies the species of each migratory bird target contained in each target area, and then precisely calculates the total number of each migratory bird species in the target habitat. Specifically, the method first monitors migratory bird monitoring images of the target habitat in real time and divides these images into multiple smaller images, ensuring that adjacent smaller images overlap. Then, the smaller images are used for target area identification, and the identified target areas are initially marked. Next, it is determined whether any overlapping areas contain duplicate target areas. If so, only one target area is retained. The image to be identified after removing duplicate target areas is then input into a fine-grained migratory bird identification model, thereby identifying the species of each migratory bird target contained in each target area, and thus accurately obtaining the total number of each migratory bird species.

[0077] Please see Figure 4 The diagram shows a structural schematic of a group identification system for monitoring migratory bird habitats according to a second embodiment of the present invention. The system includes:

[0078] Image segmentation module 10 is used to acquire migratory bird monitoring images of the target habitat at a first preset time interval, and segment the migratory bird monitoring images into a preset number of small images, wherein any two adjacent small images have overlapping parts;

[0079] Furthermore, the image segmentation module 10 also includes:

[0080] The cutting line acquisition unit is used to acquire the size information of the migratory bird monitoring image, and acquire the position coordinates of all first cutting lines according to the size information and a first preset spacing value, and acquire the position coordinates of all second cutting lines according to the position coordinates of the first cutting lines and a second preset value, wherein the first cutting lines and the second cutting lines constitute multiple overlapping areas;

[0081] The cutting execution unit is used to segment the migratory bird monitoring image into multiple smaller images based on the position coordinates of all the first cutting lines and all the position coordinates of the second cutting lines.

[0082] The initial labeling module 20 is used to input the small image into the trained migratory bird target recognition model, so as to perform initial labeling on each target region in the small image according to the target recognition result, and each target region includes at least one migratory bird target;

[0083] Furthermore, the initial marking module 20 also includes:

[0084] The historical migratory bird image acquisition unit is used to acquire multiple historical migratory bird images and, based on the historical migratory bird images, acquire the location information of all known target areas contained in each historical migratory bird image;

[0085] The historical migratory bird image annotation unit is used to annotate each historical migratory bird image according to the location information of all known target areas contained in each historical migratory bird image, and to train the migratory bird target recognition model based on the annotated historical migratory bird images.

[0086] The duplicate region detection module 30 is used to obtain the contour coordinates of the overlapping region between any two adjacent small images, and to determine whether there is a duplicate target region in the overlapping region based on the contour coordinates of the overlapping region.

[0087] Furthermore, the repeating region detection module 30 also includes:

[0088] The contour coordinate acquisition unit is used to acquire the contour coordinates of each target region marked for the first time, and to traverse all target regions based on the contour coordinates of the overlapping region and the contour coordinates of each target region to determine whether there is a target region in the overlapping region.

[0089] The duplicate region deletion module 40 is used to delete one of the duplicate target regions contained in any two adjacent small images if there are duplicate target regions in the overlapping region, so as to obtain the image to be identified after deleting the duplicate target regions.

[0090] The migratory bird species identification module 50 is used to input the image to be identified corresponding to each of the small images into the trained migratory bird fine-grained identification model, so that the migratory bird fine-grained identification model can identify each first-marked target region in the image to be identified, and obtain the species corresponding to each migratory bird target included in each target region;

[0091] Furthermore, the migratory bird species identification module 50 also includes:

[0092] The feature extraction unit is used to acquire images of migratory birds of known types, wherein each migratory bird target in the migratory bird image corresponds to a type of migratory bird, and to extract features from each migratory bird target in the migratory bird image to obtain head feature information, neck feature information, torso feature information and leg feature information corresponding to each migratory bird target;

[0093] The fine-grained recognition model training unit is used to train the migratory bird fine-grained recognition model based on the head feature information, neck feature information, trunk feature information, leg feature information corresponding to each migratory bird target, and the migratory bird type corresponding to each migratory bird target;

[0094] The migratory bird species identification execution unit is used by the trained fine-grained migratory bird identification model to identify each target region in the image to be identified based on the head feature information, the neck feature information, the trunk feature information, and the leg feature information.

[0095] The secondary labeling module 60 is used to perform secondary labeling on each migratory bird target included in each target area according to the species corresponding to each migratory bird target included in each target area, and to summarize the secondary labeling results in all the images to be identified to obtain the total number corresponding to each migratory bird species.

[0096] In another aspect, the present invention also proposes a storage medium having stored one or more programs thereon, which, when executed by a processor, implement the above-described method for group identification of migratory bird habitats.

[0097] In another aspect, the present invention also proposes a computer device, including a memory and a processor, wherein the memory is used to store computer programs and the processor is used to execute the computer programs stored in the memory to realize the above-mentioned method for group identification of migratory bird habitats.

[0098] Those skilled in the art will understand that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain stored, communicated, propagated, or transmitted programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0099] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0100] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0101] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0102] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for group identification in monitoring migratory bird habitats, applied to a migratory bird monitoring and management platform, characterized in that, The method includes: Every first preset time interval, migratory bird monitoring images of the target habitat are acquired, and the migratory bird monitoring images are divided into a preset number of small images, with any two adjacent small images having an overlapping portion; The small image is input into a trained migratory bird target recognition model to perform initial labeling on each target region in the small image based on the target recognition results, and each target region includes at least one migratory bird target; Obtain the contour coordinates of the overlapping region between any two adjacent small images, and determine whether there is a duplicate target region within the overlapping region based on the contour coordinates of the overlapping region; If there are duplicate target regions within the overlapping area, then delete one of the duplicate target regions contained in any two adjacent small images to obtain the image to be identified after deleting the duplicate target regions. The image to be identified, corresponding to each of the small images, is input into the trained fine-grained migratory bird recognition model, so that the fine-grained migratory bird recognition model can identify each first-marked target region in the image to be identified, and obtain the types of each migratory bird target included in each target region. Each migratory bird target in each target area is secondary-labeled according to the species corresponding to each migratory bird target included in each target area, and the secondary labeling results in all the images to be identified are summarized to obtain the total number corresponding to each migratory bird species. The step of inputting the image to be identified corresponding to each of the small images into the trained fine-grained migratory bird recognition model, so that the fine-grained migratory bird recognition model can identify each initially labeled target region in the image to be identified and obtain the species corresponding to each migratory bird target included in each target region, includes: Acquire images of migratory birds of known types, where each migratory bird target in the image corresponds to a type of migratory bird. Extract features from each migratory bird target in the image to obtain head feature information, neck feature information, torso feature information, and leg feature information corresponding to each migratory bird target. The fine-grained migratory bird recognition model is trained based on the head feature information, neck feature information, trunk feature information, leg feature information, and migratory bird type corresponding to each migratory bird target. The trained fine-grained migratory bird recognition model identifies each target region in the image to be recognized based on the head feature information, the neck feature information, the trunk feature information, and the leg feature information.

2. The method for group identification of migratory bird habitats according to claim 1, characterized in that, The step of acquiring migratory bird monitoring images of the target habitat at first preset time intervals, and dividing the migratory bird monitoring images into a preset number of smaller images, wherein any two adjacent smaller images have overlapping portions, includes: The size information of the migratory bird monitoring image is obtained, and the position coordinates of all first cutting lines are obtained according to the size information and a first preset spacing value. The position coordinates of all second cutting lines are obtained according to the position coordinates of the first cutting lines and a second preset value. The first cutting lines and the second cutting lines constitute multiple overlapping areas. The migratory bird monitoring image is segmented into multiple smaller images based on the position coordinates of all the first cutting lines and all the position coordinates of all the second cutting lines.

3. The method for group identification of migratory bird habitats according to claim 1, characterized in that, The step of inputting the small image into a trained migratory bird target recognition model, and initially labeling each target region in the small image according to the target recognition result, wherein each target region includes at least one migratory bird target, includes: Acquire multiple historical migratory bird images, and obtain the location information of all known target areas contained in each historical migratory bird image based on the historical migratory bird images; Each historical migratory bird image is labeled based on the location information of all known target areas contained in each historical migratory bird image, and the migratory bird target recognition model is trained based on the labeled historical migratory bird images.

4. The method for group identification of migratory bird habitats according to claim 2, characterized in that, The step of obtaining the contour coordinates of the overlapping region between any two adjacent small images, and determining whether there is a duplicate target region within the overlapping region based on the contour coordinates of the overlapping region, includes: Obtain the contour coordinates of each target region marked for the first time, and traverse all target regions based on the contour coordinates of the overlapping region and the contour coordinates of each target region to determine whether there is a target region within the overlapping region.

5. A group identification system for monitoring migratory bird habitats, characterized in that, The system includes: The image segmentation module is used to acquire migratory bird monitoring images of the target habitat at a first preset time interval, and to segment the migratory bird monitoring images into a preset number of small images, with any two adjacent small images having an overlapping portion. The initial labeling module is used to input the small image into the trained migratory bird target recognition model, so as to perform initial labeling on each target region in the small image according to the target recognition result, and each target region includes at least one migratory bird target; The duplicate region detection module is used to obtain the contour coordinates of the overlapping region between any two adjacent small images, and to determine whether there is a duplicate target region in the overlapping region based on the contour coordinates of the overlapping region. The duplicate region deletion module is used to delete one of the duplicate target regions contained in any two adjacent small images if there are duplicate target regions in the overlapping region, so as to obtain the image to be identified after deleting the duplicate target regions. The migratory bird species identification module is used to input the image to be identified corresponding to each of the small images into the trained migratory bird fine-grained identification model, so that the migratory bird fine-grained identification model can identify each first-marked target region in the image to be identified, and obtain the species corresponding to each migratory bird target included in each target region; The secondary labeling module is used to perform secondary labeling on each migratory bird target included in each target region according to the species corresponding to each migratory bird target included in each target region, and to summarize the secondary labeling results in all the images to be identified to obtain the total number corresponding to each migratory bird species. The migratory bird species identification module also includes: The feature extraction unit is used to acquire images of migratory birds of known types, wherein each migratory bird target in the migratory bird image corresponds to a type of migratory bird, and to extract features from each migratory bird target in the migratory bird image to obtain head feature information, neck feature information, torso feature information and leg feature information corresponding to each migratory bird target; The fine-grained recognition model training unit is used to train the fine-grained recognition model of migratory birds based on the head feature information, neck feature information, trunk feature information, leg feature information corresponding to each migratory bird target, and the migratory bird type corresponding to each migratory bird target; The migratory bird species identification execution unit is used by the trained fine-grained migratory bird identification model to identify each target region in the image to be identified based on the head feature information, the neck feature information, the torso feature information, and the leg feature information.

6. The population identification system for monitoring migratory bird habitats according to claim 5, characterized in that, The image segmentation module further includes: The cutting line acquisition unit is used to acquire the size information of the migratory bird monitoring image, and acquire the position coordinates of all first cutting lines according to the size information and a first preset spacing value, and acquire the position coordinates of all second cutting lines according to the position coordinates of the first cutting lines and a second preset value, wherein the first cutting lines and the second cutting lines constitute multiple overlapping areas; The cutting execution unit is used to segment the migratory bird monitoring image into multiple smaller images based on the position coordinates of all the first cutting lines and all the position coordinates of the second cutting lines.

7. The population identification system for monitoring migratory bird habitats according to claim 5, characterized in that, The initial marking module also includes: The historical migratory bird image acquisition unit is used to acquire multiple historical migratory bird images and, based on the historical migratory bird images, acquire the location information of all known target areas contained in each historical migratory bird image; The historical migratory bird image annotation unit is used to annotate each historical migratory bird image according to the location information of all known target areas contained in each historical migratory bird image, and to train the migratory bird target recognition model based on the annotated historical migratory bird images.

8. The population identification system for monitoring migratory bird habitats according to claim 6, characterized in that, The repeating region detection module further includes: The contour coordinate acquisition unit is used to acquire the contour coordinates of each target region marked for the first time, and to traverse all target regions according to the contour coordinates of the overlapping region and the contour coordinates of each target region to determine whether there is a target region in the overlapping region.

Images