Goose egg laying behavior accurate monitoring method, device and system and storage medium

Abstract

The invention discloses a goose egg laying behavior accurate monitoring method, device and system and a storage medium, and the method comprises the steps: obtaining a goose house monitoring video, and sleeving a goose neck part in a goose house with identification neck rings with different fluorescent patterns; performing framing processing on the monitoring video of each goose house to obtain a single goose house image; preprocessing the acquired goose house image; marking a neck ring and marking spatio-temporal information of the goose eggs on the preprocessed image; according to the marked image, an improved YOLOvX target detection model is adopted for training and optimization, and an identity recognition model of geese in the goose house is obtained; detecting an input image of a to-be-monitored goose with a neck ring by using the model to obtain identity information of the corresponding goose and time and space position information of goose egg production; the goose identification neck ring data and the spatio-temporal information of goose egg laying are subjected to post-processing analysis, and accurate identification of goose egg laying in the goose house is achieved. The method can solve the problem that manual recognition statistics wastes time and labor, and provides powerful guarantee for precise breeding in the goose industry.

Description

Technical Field

[0001] The present invention relates to a method, device, system and storage medium for accurately monitoring goose egg-laying behavior, belonging to the technical field of intelligent poultry breeding of livestock and poultry. Background Art

[0002] As consumers' requirements for the quality and safety of poultry products continue to increase, the breeding industry faces the dual pressure of improving breeding efficiency and product quality. Breeders need to adopt efficient breeding strategies to ensure the high quality of poultry growth while reducing breeding costs. In terms of poultry selection and breeding, the egg-laying rate of breeding geese can be used to analyze their egg-laying ability and subsequent improvement directions. However, manually counting the egg-laying rate of breeding geese in the goose house and identifying egg-laying behavior is a time-consuming and laborious task. At the same time, due to the lack of light at night, the identification of the markers is blurred, and the visual fatigue caused by manually dragging the monitoring playback will make the accuracy of manual counting impossible to guarantee. How to reduce the burden of goose egg statistics and improve the accuracy of egg-laying identification has become a crucial issue. Therefore, the industry urgently needs to develop an automated method and system for accurately monitoring goose egg-laying behavior. Summary of the invention

[0003] In view of this, the present invention provides a method, device, system and storage medium for accurately monitoring goose egg-laying behavior, which can solve the problem of time-consuming and labor-intensive manual identification and statistics, and can effectively and accurately identify which goose lays which egg, find out the correspondence between geese and goose eggs, provide strong guarantees for accurate breeding in the goose industry, help simplify the breeding process and improve livestock and poultry farming practices.

[0004] The first object of the present invention is to provide a method for accurately monitoring goose egg-laying behavior.

[0005] The second object of the present invention is to provide a precise monitoring system for goose egg-laying behavior.

[0006] The third object of the present invention is to provide an inspection robot.

[0007] A fourth object of the present invention is to provide a computer-readable storage medium.

[0008] The first object of the present invention can be achieved by adopting the following technical solutions:

[0009] A method for accurately monitoring goose egg-laying behavior, the method comprising:

[0010] Obtaining surveillance video of a goose house, wherein the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns;

[0011] Perform frame processing on the surveillance video of each goose house to obtain a single goose house image;

[0012] Preprocessing the acquired goose house images;

[0013] The pre-processed images are labeled with neck rings and the temporal and spatial information of the goose eggs;

[0014] According to the annotated images, the improved YOLOvX target detection model is used for training and optimization to obtain the goose monitoring detection and identification model in the goose house;

[0015] The goose identification model is used to detect the input image of the goose with a neck ring to be monitored, and the identity information of the corresponding goose and the time and space location information of goose egg production are obtained;

[0016] According to the temporal and spatial correlation between geese and their eggs, post-processing analysis is performed on the goose neck ring data and the spatiotemporal information of goose egg laying to achieve accurate identification of geese laying eggs in the goose house.

[0017] Furthermore, the preprocessing of the acquired goose house image specifically includes:

[0018] The acquired goose house image is enhanced by using image enhancement technology;

[0019] The enhanced goose house image is sliced ​​at a preset resolution and repetition rate to obtain slice images of uniform specifications.

[0020] Furthermore, the labeling of the neck ring and the spatiotemporal information of the goose eggs produced on the pre-processed image specifically includes:

[0021] The spatiotemporal information of the neck ring and the goose eggs produced in the preprocessed image is annotated using annotation frames of different colors. The fluorescent pattern of the neck ring is a rectangle, a circle, a triangle and a combination thereof.

[0022] Furthermore, the improved YOLOvX target detection model introduces the EMA attention mechanism, replaces the FasterNet lightweight model backbone on the Backbone side, and finally adopts the GioU loss function, with 200 training rounds, 16 training batches, and an initial learning rate of 1×10 -4 , decay is 5×10 -4 , using the SGD optimizer.

[0023] Furthermore, the goose identity recognition model is used to detect the input image of the goose to be monitored with the neck ring, and the identity information of the corresponding goose and the time and space location information of the goose egg production are obtained, which specifically includes:

[0024] For the input image of the goose with the collar, a series of convolutional layers in the goose identification model extract image features;

[0025] The image features are processed by a specific detection head, which uses anchor points to predict the identity of the goose, as well as the time and spatial location of the goose egg production;

[0026] Non-maximum suppression is used for each predicted neck ring and goose egg to reduce overlapping detection results.

[0027] Furthermore, according to the temporal and spatial correlation of the goose and the goose eggs, the goose identification neck ring data and the temporal and spatial information of goose egg laying are post-processed and analyzed, specifically including:

[0028] According to the temporal and spatial correlation between geese and their eggs, statistical analysis is performed on the goose neck ring data and the temporal and spatial information of goose egg laying, generating a detailed report including the maternal matching of goose eggs, the number of eggs laid and the nesting time.

[0029] Furthermore, the maternal line matching of goose eggs refers to matching each identified goose egg with the nearest identification neck ring pattern, so as to obtain the maternal breeding goose identity information of the goose egg.

[0030] The second object of the present invention can be achieved by adopting the following technical solutions:

[0031] A device for accurately monitoring goose egg-laying behavior, the device comprising:

[0032] An acquisition unit, used for acquiring surveillance video of a goose house, wherein the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns;

[0033] A frame processing unit, used for performing frame processing on each goose house monitoring video to obtain a single goose house image;

[0034] A preprocessing unit, used for preprocessing the acquired goose house image;

[0035] A labeling unit, used to label the pre-processed image with neck rings and spatiotemporal information of the goose eggs;

[0036] A training and optimization unit, used for training and optimizing an improved YOLOvX target detection model according to the labeled images to obtain a goose identity recognition model in a goose house;

[0037] A detection unit, used to detect the input image of the goose to be monitored with the neck ring using the goose identity recognition model, and obtain the identity information of the corresponding goose and the time and space location information of the goose egg production;

[0038] The evaluation unit is used to perform post-processing analysis on the goose neck ring data and the spatiotemporal information of goose egg laying according to the temporal and spatial correlation of the geese and the eggs they lay, so as to achieve accurate identification of geese laying eggs in the goose house.

[0039] The third object of the present invention can be achieved by adopting the following technical solutions:

[0040] A precise monitoring system for goose egg-laying behavior, the system comprising a local industrial computer, a server, a switch and a plurality of cameras, the cameras are arranged in a goose house, the cameras are connected to the local industrial computer through the switch, and the local industrial computer is connected to the server;

[0041] The local industrial computer is used to obtain surveillance videos of a goose house, where the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns; perform frame processing on the surveillance videos of each goose house to obtain a single goose house image; pre-process the obtained goose house image; annotate the pre-processed image with identification neck rings and spatiotemporal information of the goose eggs produced; upload the annotated image to the server; use the goose identity recognition model issued by the server to detect the input image of the goose to be monitored with the neck ring, and obtain the identity information of the corresponding goose and the time and space location information of the goose egg production; according to the time and space correlation between the goose and the goose eggs produced, post-process and analyze the goose identification neck ring data and the spatiotemporal information of the goose egg production, so as to realize the accurate identification of the goose egg production in the goose house;

[0042] The server is used to train and optimize the improved YOLOvX target detection model according to the labeled images, obtain the goose identity recognition model in the goose house, and deploy the goose identity recognition model on the local industrial computer.

[0043] The fourth object of the present invention can be achieved by adopting the following technical solutions:

[0044] A computer-readable storage medium stores a program, and when the program is executed by a processor, the above-mentioned method for accurately monitoring goose egg-laying behavior is implemented.

[0045] The present invention has the following beneficial effects compared with the prior art:

[0046] The present invention improves the efficiency and accuracy of goose egg maternal lineage identification. Compared with the traditional manual method, goose eggs can be automatically identified and analyzed, which significantly reduces labor intensity and improves the speed of data processing; in addition, the present invention also ensures the reliability and accuracy of the analysis results through optimized image processing and data analysis, thereby helping breeders to better understand the egg-laying status of breeding geese, thereby improving livestock and poultry breeding practices and increasing production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in 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 described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0048] Figure 1 This is a structural block diagram of the precise monitoring system for goose egg-laying behavior according to Example 1 of the present invention.

[0049] Figure 2 This is a flow chart of the method for accurately monitoring goose egg-laying behavior according to Example 1 of the present invention.

[0050] Figure 3 This is a structural diagram of the goose identity recognition model of the improved YOLOvX according to Example 1 of the present invention.

[0051] Figure 4 This is a schematic diagram of the interaction of the user interface of Example 1 of the present invention.

[0052] Figure 5 This is a structural block diagram of a device for accurately monitoring goose egg-laying behavior according to Example 2 of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0054] Embodiment 1:

[0055] like Figure 1 As shown, this embodiment provides a precise monitoring system for goose egg-laying behavior, which can effectively improve the accuracy and efficiency of goose egg recognition and is suitable for automated analysis of individual goose egg laying. The system includes a local industrial computer 101, a server 102, a switch 103 and eight cameras 104. The camera 104 is arranged in the goose house. The camera 104 is connected to the local industrial computer 101 through the switch 103, and the local industrial computer 101 is connected to the server 102. The camera 104 adopts a Hikvision camera and the server 102 adopts a GPU server.

[0056] like Figure 2 As shown, this embodiment also provides a method for accurately monitoring goose egg-laying behavior, which is implemented by the above-mentioned accurate monitoring system for goose egg-laying behavior, and includes the following steps:

[0057] S201, obtaining surveillance video of the goose house.

[0058] Step S201 of this embodiment is implemented by the local industrial computer 101. The camera 104 and the switch 103 are connected by a network cable, and then the local industrial computer 101 and the switch 103 are connected by a network cable to obtain the IP addresses of different cameras, and put neck rings with different patterns on different goose necks. The shapes of the neck rings include grid cylinders, semi-cones or back fluorescent plates. The patterns of the neck rings are rectangles, circles, triangles and combinations thereof. Real-time goose house monitoring video is obtained through RTSP streaming.

[0059] S202: Perform frame processing on the surveillance video of each goose house to obtain a single goose house image.

[0060] Step S202 of this embodiment is implemented by the local industrial computer 101. After the goose house monitoring video is obtained, the obtained goose house monitoring video is frame-processed to obtain the goose house image, and the image is saved.

[0061] S203, pre-processing the acquired goose house image.

[0062] Step S203 of this embodiment is implemented by the local industrial computer 101, and the image is enhanced by image enhancement techniques such as adaptive histogram equalization and high-pass filtering; the enhanced goose house image is sliced ​​at a preset resolution and repetition rate to obtain sliced ​​images of uniform specifications, wherein the obtained goose house image has a resolution of 2304*1296, the preset resolution is uniformly 640*640, and the preset repetition rate is 0.2, thereby ensuring the consistency and efficiency of subsequent processing.

[0063] S204, marking the neck ring and the spatiotemporal information of the goose eggs produced in the pre-processed image.

[0064] Step S204 of this embodiment is implemented by the local industrial computer 101. The identification neck ring and the spatiotemporal information of the goose eggs in the pre-processed image are marked with annotation boxes of different colors to facilitate visual recognition and subsequent analysis, and the marked images of the geese in the goose house are uploaded to the server 102.

[0065] S205. According to the labeled images, an improved YOLOvX target detection model is used for training and optimization to obtain a goose identity recognition model in a goose house.

[0066] Step S205 of this embodiment is implemented by the server 102, which divides the labeled images into a training set and a test set in a ratio of 9:1. The training set data is used to further train and optimize the improved YOLOvX (X can be 5 / 6 / 7 / 8 / 9 / 10, etc.) target detection model to ensure that the model can achieve high-precision identification of neck rings and goose eggs on an unseen test set, and finally obtain a goose identity recognition model in a goose house. The structure of the improved YOLOvX target detection model is as follows: Figure 3 As shown in the figure, it introduces the EMA attention mechanism, replaces the FasterNet lightweight model backbone on the Backbone side, and finally adopts the GioU loss function. The number of training rounds is 200, the training batch size is 16, and the initial learning rate is 1×10 -4 , decay is 5×10 -4 , using the SGD optimizer.

[0067] During the training process, the model learns to recognize different types of identification neck rings and goose eggs. The shapes of the neck rings include grid cylinders, semi-cones or back fluorescent plates. The fluorescent patterns of the identification neck rings are rectangles, circles, triangles and their combinations, and can accurately distinguish them from the background. After the training is completed, the obtained goose identity recognition model is converted for deployment based on the ONNX format, and the algorithm using the model is redesigned to lower the equipment and computing power threshold for using this method.

[0068] S206, using the goose identity recognition model to detect the input image of the goose to be monitored with the neck ring, and obtain the identity information of the corresponding goose and the time and space location information of the goose egg production.

[0069] Step S206 of this embodiment is implemented by the local industrial computer 101. After the goose identity recognition model is deployed on the local industrial computer 101, the goose identity recognition model is used to detect the input image of the goose to be monitored with the neck ring, and the identity information of the corresponding goose and the time and space location information of the goose egg production are obtained, which specifically includes:

[0070] S2061. For the input image of the goose to be monitored with a neck ring, extract image features through a series of convolutional layers in the goose identification model.

[0071] S2062. Processing image features through a specific detection head, which uses anchor points to predict the identity information of the goose, as well as the world and spatial position information of the goose eggs produced in the goose house.

[0072] S2063. Use non-maximum suppression (NMS) for each predicted neck ring and goose egg to reduce overlapping detection results, ensuring that each identified neck ring and goose egg is detected only once.

[0073] S207. Based on the temporal and spatial correlation between the geese and the eggs they lay, post-process and analyze the goose neck ring data and the temporal and spatial information of the goose egg laying to achieve accurate identification of the geese laying eggs in the goose house.

[0074] Step S206 of this embodiment is implemented by the local industrial computer 101. According to the time and space correlation between the geese and the eggs they produce, the goose identification neck ring data (the identity information of the goose) and the goose egg data (the time and space location information of the goose eggs produced in the goose house) are statistically analyzed to generate a detailed report including the maternal matching of the goose eggs, the number of eggs laid and the nesting time. Among them, the maternal matching of the goose eggs refers to matching each identified goose egg with the nearest identification neck ring pattern, so as to obtain the identity information of the maternal breeding goose of the goose egg. These data are used to realize the accurate identification of the egg laying of breeding geese. By analyzing the number of eggs laid and the nesting time of breeding geese with different identification neck ring patterns, breeding geese with excellent egg laying rate can be predicted at an early stage.

[0075] In order to ensure that the accurate monitoring method of goose egg-laying behavior in this embodiment is easy to operate and efficient and practical, a QT-based user interface is designed. Figure 4 As shown, the main components of the interface and their functions are described in detail below:

[0076] The data input and function operation module can receive the recognized video push streams from eight independent cameras; the administrator can choose the display mode of single or multiple camera views through the interface, including arranging and displaying the views of all eight cameras at the same time or selecting the video of any camera for enlarged display.

[0077] System recognition result display module, this area is specifically used to display the results of goose egg detection. Each identified goose egg is matched with the nearest neck ring pattern, and will be displayed in a rectangular frame of the corresponding color of the neck ring, and the corresponding neck ring text category will be displayed above, so that users can quickly identify the mother goose to which the goose egg belongs.

[0078] The data statistics table display module provides a detailed data statistics table display area to record the event flow in real time. This embodiment lists the specific number of eggs laid by the breeding geese in each goose house and their nesting time.

[0079] It should be noted that although the above method operations are described in a particular order in the accompanying drawings, this does not require or imply that these operations must be performed in this particular order, or that all of the operations shown must be performed to achieve the desired results. On the contrary, the steps depicted may be performed in a different order. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0080] Embodiment 2:

[0081] like Figure 5 As shown, this embodiment provides a device for accurately monitoring goose egg-laying behavior, which includes an acquisition unit 501, a frame processing unit 502, a preprocessing unit 503, a labeling unit 504, a training and optimization unit 505, a detection unit 506 and an evaluation unit 507. The specific functions of each unit are as follows:

[0082] An acquisition unit 501 is used to acquire a surveillance video of a goose house, wherein the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns;

[0083] A frame processing unit 502 is used to perform frame processing on the monitoring video of each goose house to obtain a single goose house image;

[0084] A preprocessing unit 503 is used to preprocess the acquired goose house image;

[0085] A labeling unit 504 is used to label the pre-processed image with neck rings and spatiotemporal information of the goose eggs;

[0086] The training and optimization unit 505 is used to train and optimize the geese in the goose house using the improved YOLOvX target detection model according to the labeled images, so as to obtain a goose identity recognition model;

[0087] A detection unit 506 is used to detect the input image of the goose with the neck ring to be monitored using the goose identity recognition model to obtain the identity information of the corresponding goose and the time and space location information of goose egg production;

[0088] The evaluation unit 507 is used to perform post-processing analysis on the goose identification neck ring data and the spatiotemporal information of goose egg laying according to the temporal and spatial correlation between the goose and the goose eggs, so as to realize accurate identification of goose eggs laid in the goose house.

[0089] The specific implementation of each unit in this embodiment can refer to the above-mentioned embodiment 1, and will not be described one by one here; it should be noted that the system provided in this embodiment is only illustrated by the division of the above-mentioned functional units. In actual applications, the above-mentioned functions can be assigned to different functional units as needed, that is, the internal structure can be divided into different functional units to complete all or part of the functions described above.

[0090] Embodiment 3:

[0091] This embodiment provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for accurately monitoring goose egg-laying behavior of the above-mentioned embodiment 1 is implemented, which is specifically as follows:

[0092] A surveillance video of a goose house is obtained, in which the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns; the surveillance video of each goose house is framed to obtain a single goose house image; the obtained goose house image is preprocessed; the preprocessed image is annotated with the identification neck ring and the spatiotemporal information of the goose eggs produced; based on the annotated image, an improved YOLOvX target detection model is used for training and optimization to obtain a goose identity recognition model in the goose house; the goose identity recognition model is used to detect the input image of the goose to be monitored with the neck ring, and the identity information of the corresponding goose and the time and spatial location information of the goose egg production are obtained; based on the time and space correlation between the goose and the goose eggs produced, the goose identification neck ring data and the spatiotemporal information of the goose egg production are post-processed and analyzed to achieve accurate identification of the goose egg production in the goose house.

[0093] It should be noted that the computer-readable storage medium of the present embodiment may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0094] In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, device or device. In this embodiment, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium may also be any computer-readable storage medium other than a computer-readable storage medium, which may send, propagate, or transmit a program used by or in combination with an instruction execution system, device or device. The computer program contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0095] The computer readable storage medium can be written in one or more programming languages ​​or a combination thereof to execute the computer program of the present embodiment, and the programming language includes an object-oriented programming language, such as Java, Python, C++, and a conventional procedural programming language, such as C or a similar programming language. The program can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect through the Internet).

[0096] To sum up, the present invention effectively overcomes many shortcomings and deficiencies of the prior art. By designing different neck ring patterns as breeding goose identity identification, different geese in the same goose house can be effectively distinguished. By adopting advanced deep learning technology, the neck ring identity identification patterns and goose eggs of breeding geese can be automatically identified, which significantly improves the speed and accuracy of statistics. Compared with traditional manual methods, the present invention reduces dependence on manpower, improves the degree of automation of data processing, and ensures the reliability and repeatability of the results.

[0097] The above is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solution and inventive concept of the present invention within the scope disclosed by the present invention, which shall fall within the protection scope of the present invention.

Claims

1. A method for accurately monitoring goose egg-laying behavior, characterized in that: The method comprises: Obtaining surveillance video of a goose house, wherein the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns; Perform frame processing on the surveillance video of each goose house to obtain a single goose house image; Preprocessing the acquired goose house images; The pre-processed images are labeled with neck rings and the temporal and spatial information of the goose eggs; According to the annotated images, the improved YOLOvX target detection model is used for training and optimization to obtain the goose identity recognition model in the goose house; The goose identification model is used to detect the input image of the goose with a neck ring to be monitored, and the identity information of the corresponding goose and the time and space location information of goose egg production are obtained; According to the temporal and spatial correlation between geese and their eggs, post-processing analysis is performed on the goose neck ring data and the spatiotemporal information of goose egg laying to achieve accurate identification of geese laying eggs in the goose house.

2. The method for accurately monitoring goose egg-laying behavior according to claim 1, characterized in that: The preprocessing of the acquired single goose house image specifically includes: The acquired goose house image is enhanced by using image enhancement technology; The enhanced single goose house image is sliced ​​at a preset resolution and repetition rate to obtain slice images of uniform specifications.

3. The method for accurately monitoring goose egg-laying behavior according to claim 1, characterized in that: The labeling of the neck ring and the spatiotemporal information of the goose eggs produced on the pre-processed image specifically includes: The spatiotemporal information of the neck ring and the goose eggs produced in the preprocessed image is annotated using annotation frames of different colors. The fluorescent pattern of the neck ring is a rectangle, a circle, a triangle and a combination thereof.

4. The method for accurately monitoring goose egg-laying behavior according to claim 1, characterized in that: The improved YOLOvX target detection model introduces the EMA attention mechanism, replaces the FasterNet lightweight model backbone on the Backbone side, and finally adopts the GioU loss function. The training rounds are 200, the training batch size is 16, and the initial learning rate is 1×10 -4 , decay is 5×10 -4 , using the SGD optimizer.

5. The method for accurately monitoring goose egg-laying behavior according to claim 1, characterized in that: The goose identification model is used to detect the input image of the goose to be monitored with the neck ring, and the identity information of the corresponding goose and the time and space location information of the goose egg production are obtained, which specifically includes: For the input image of the goose with the collar, a series of convolutional layers in the goose identification model extract image features; The image features are processed by a specific detection head, which uses anchor points to predict the identity of the goose, as well as the time and spatial location of the goose egg production; Non-maximum suppression is used for each predicted neck ring and goose egg to reduce overlapping detection results.

6. The method for accurately monitoring goose egg-laying behavior according to claim 1, characterized in that: According to the temporal and spatial correlation between the goose and the goose eggs, the goose identification neck ring data and the temporal and spatial information of goose egg laying are post-processed and analyzed, specifically including: According to the temporal and spatial correlation between geese and their eggs, statistical analysis is performed on the goose neck ring data and the temporal and spatial information of goose egg laying, generating a detailed report including the maternal matching of goose eggs, the number of eggs laid and the nesting time.

7. The method for accurately monitoring goose egg-laying behavior according to claim 6, characterized in that: The maternal line matching of goose eggs refers to matching each identified goose egg with the nearest identification neck ring pattern, so as to obtain the maternal line breeding goose identity information of the goose egg.

8. A precise monitoring device for goose egg-laying behavior, characterized in that: The device comprises: An acquisition unit, used for acquiring surveillance video of a goose house, wherein the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns; A frame processing unit, used for performing frame processing on each goose house monitoring video to obtain a single goose house image; A preprocessing unit, used for preprocessing the acquired goose house image; A labeling unit, used to label the pre-processed image with neck rings and spatiotemporal information of the goose eggs; A training and optimization unit, used for training and optimizing an improved YOLOvX target detection model according to the labeled images to obtain a goose identity recognition model in a goose house; A detection unit, used to detect the input image of the goose with the neck ring to be monitored by using the goose identity recognition model, and obtain the identity information of the corresponding goose and the time and space location information of goose egg production; The evaluation unit is used to perform post-processing analysis on the goose neck ring data and the spatiotemporal information of goose egg laying according to the temporal and spatial correlation of the geese and the eggs they lay, so as to achieve accurate identification of geese laying eggs in the goose house.

9. A precise monitoring system for goose egg-laying behavior, characterized in that: The system includes a local industrial computer, a server, a switch and a plurality of cameras, wherein the cameras are arranged in a goose house, the cameras are connected to the local industrial computer through the switch, and the local industrial computer is connected to the server; The local industrial computer is used to obtain surveillance videos of a goose house, where the necks of the geese in the goose house are covered with identification neck rings with different fluorescent patterns; perform frame processing on the surveillance videos of each goose house to obtain a single goose house image; and perform preprocessing on the obtained goose house image; The pre-processed images are labeled with neck rings and the temporal and spatial information of the goose eggs; Upload the annotated image to the server; The goose identification model sent by the server is used to detect the input image of the goose with the neck ring to be monitored, and the identity information of the corresponding goose and the time and space location information of goose egg production are obtained; according to the time and space correlation between the goose and the goose eggs, the goose identification neck ring data and the spatiotemporal information of goose egg production are post-processed and analyzed to achieve accurate identification of goose egg production in the goose house; The server is used to train and optimize the improved YOLOvX target detection model according to the labeled images, obtain the goose identity recognition model in the goose house, and deploy the goose identity recognition model on the local industrial computer.

10. A computer-readable storage medium storing a program, characterized in that: When the program is executed by the processor, the method for accurately monitoring goose egg-laying behavior as described in any one of claims 1 to 7 is implemented.

Images