A fish body positioning and tracking device and method

Abstract

The embodiment of the application provides a fish body positioning and tracking device and method, an RFID card reading coil is used for detecting a label ID carried by a target fish body; the label ID and a corresponding timestamp are sent to the control module; the control module is used for receiving and responding to the label ID and the corresponding timestamp, controlling the camera to collect an image frame of the target fish body; through a preset fish body target detection model, the fish body is detected according to the image frame of the target fish body; visual ID is set according to the detected fish body information; the label ID and the visual ID are bound; through the visual ID and a preset multi-target tracking algorithm, the camera is controlled to track and monitor the target fish body. Through the scheme of the embodiment of the application, the fish body can be continuously tracked and monitored through the visual ID, and the problems of low efficiency and high cost in the fish body positioning and tracking process are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, in particular to a fish body positioning and tracking device and method. BACKGROUND

[0002] In the process of aquaculture, accurately collecting the feeding behavior and feeding amount information of each fish in the fish group is of great significance for optimizing feed formula, feeding strategy, improving feed utilization rate, and cultivating excellent strains that are resistant to coarse feed and save feed (not dependent on fish meal and other high-quality protein sources, high feed conversion rate). Traditional methods, such as feeding fish one by one through physical isolation or limiting single fish to enter the feeding area, can achieve accurate collection of individual information, but they greatly deviate from the characteristics of group gathering and food competition in natural breeding environment, and have the problems of low efficiency and high cost. SUMMARY

[0003] The purpose of the embodiments of the present application is to provide a fish body positioning and tracking device and method to solve the problems of low efficiency and high cost in the process of fish body positioning and tracking. The specific technical solutions are as follows:

[0004] The first aspect of the embodiments of the present application first provides a fish body positioning and tracking device, which comprises:

[0005] a background plate arranged at the bottom, an RFID card reading coil arranged at the side wall, and a camera arranged at the top, wherein the shooting range of the camera includes the area where the background plate and the RFID card reading coil are located; the RFID card reading coil and the camera are in communication connection with the control module; the device includes an open entrance and an exit;

[0006] The RFID card reading coil is used to detect the tag ID carried by the target fish body, and send the tag ID and the corresponding time stamp to the control module.

[0007] The control module is used to receive and respond to the tag ID and the corresponding time stamp, control the camera to collect the image frame of the target fish body, perform fish body detection according to the image frame of the target fish body through a preset fish body target detection model, set a visual ID according to the detected fish body information, bind the tag ID and the visual ID, and control the camera to track and monitor the target fish body through the visual ID and a preset multi-target tracking algorithm.

[0008] In a possible implementation, the control module is specifically configured to calculate distances between each fish body and the RFID card reading coil when the camera captures an image frame containing multiple fish bodies, by using a preset fish body position recognition model; according to the calculated distances, a fish body closest to the RFID card reading coil is set as a target fish body, and the visual ID is set for the target fish body.

[0009] In a possible implementation, the control module is further configured to unbind the tag ID and the visual ID when the target fish body is detected by the camera to be out of the shooting range of the camera.

[0010] In a possible implementation, the control module is specifically configured to capture images of the target fish body at multiple time points by using the camera; and to identify behavior data of the target fish body according to the captured images at the multiple time points, wherein the behavior data includes a visual ID, a position, and a behavior track.

[0011] In a possible implementation, the device is a frame structure.

[0012] The background plate is arranged at the bottom of the frame structure, the RFID card reading coil is arranged at the side wall of the frame structure, the camera is arranged at the top of the frame structure, and the frame structure includes an open entrance and an open exit.

[0013] A second aspect of the embodiment of the application provides a fish body positioning and tracking method applied to a control module of a fish body positioning and tracking device, and the method includes the following steps.

[0014] Receiving and responding to the tag ID and the corresponding timestamp, the camera is controlled to capture an image frame of the target fish body.

[0015] Detecting a fish body according to the image frame of the target fish body by using a preset fish body target detection model.

[0016] Setting a visual ID according to the detected fish body information.

[0017] Binding the tag ID and the visual ID.

[0018] Controlling the camera to track and monitor the target fish body by using the visual ID and a preset multi-target tracking algorithm.

[0019] In a possible implementation, the step of setting a visual ID for the target fish body according to the image frame of the target fish body includes the following steps.

[0020] When the camera captures an image frame containing multiple fish bodies, the distance between each fish body and the RFID card reading coil is calculated by a preset fish body position recognition model;

[0021] According to the calculated distance, the fish body closest to the RFID card reading coil is taken as a target fish body, and the visual ID is set for the target fish body.

[0022] In a possible implementation, the method further includes:

[0023] When the target fish body is detected by the camera and leaves the shooting range of the camera, the binding between the tag ID and the visual ID is released.

[0024] In a possible implementation, the control module is specifically configured to: acquire images of the target fish body at multiple time points by the camera; and identify behavior data of the target fish body according to the acquired images at the multiple time points, wherein the behavior data includes a visual ID, a position, and a behavior track.

[0025] Another aspect of the embodiment of the present application further provides an electronic device, which includes:

[0026] a memory configured to store a computer program;

[0027] a processor configured to execute the program stored on the memory, so as to implement any of the fish body positioning and tracking methods.

[0028] Another aspect of the embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement any of the fish body positioning and tracking methods.

[0029] Another aspect of the embodiment of the present application further provides a computer program product containing instructions, which, when running on a computer, causes the computer to execute any of the fish body positioning and tracking methods.

[0030] The embodiment of the present application has the following beneficial effects:

[0031] The embodiment of the present application provides a fish body positioning and tracking device and method, the device comprises: a background plate arranged at the bottom; an RFID card reading coil arranged at the side wall; a camera arranged at the top, the shooting range of the camera comprises the area where the background plate and the RFID card reading coil are located; the RFID card reading coil and the camera are in communication connection with the control module; the device comprises an open entrance and an exit; the RFID card reading coil is used for detecting the label ID carried by the target fish body; the label ID and the corresponding timestamp are sent to the control module; the control module is used for receiving and responding to the label ID and the corresponding timestamp, and controlling the camera to collect the image frame of the target fish body; through a preset fish body target detection model, the fish body is detected according to the image frame of the target fish body; visual ID is set according to the detected fish body information; the label ID and the visual ID are bound; through the visual ID and a preset multi-target tracking algorithm, the camera is controlled to track and monitor the target fish body. Through the scheme of the embodiment of the present application, only the background plate, the RFID card reading coil and the camera are needed, the label ID of the fish body is read through the RFID card reading coil, then the visual ID of the fish body is set through the camera, so that the label ID and the visual ID are bound, the fish body is continuously tracked and monitored through the visual ID, and the problems of low efficiency and high cost in the fish body positioning and tracking process are solved.

[0032] Of course, implementing any product or method of the present application does not necessarily require achieving all the advantages described above at the same time. BRIEF DESCRIPTION OF DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other embodiments can also be obtained by those skilled in the art according to these drawings.

[0034] Figure 1 Another structural schematic view of the fish body positioning and tracking device provided by the embodiment of the present application;

[0035] Figure 2 A flowchart of the fish body positioning and tracking method corresponding to the embodiment of the present application. DETAILED DESCRIPTION

[0036] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art based on the present application belong to the scope of protection of the present application.

[0037] In a first aspect of the embodiments of the present application, a fish body positioning and tracking device is provided, which comprises:

[0038] a background plate arranged at the bottom, an RFID (Radio Frequency Identification) card reading coil arranged at the side wall, and a camera arranged at the top, wherein the shooting range of the camera includes the area where the background plate and the RFID card reading coil are located; the RFID card reading coil is in communication connection with the camera and the control module; the device comprises an open entrance and an exit;

[0039] The RFID card reading coil is configured to detect a tag ID carried by a target fish body and send the tag ID and a corresponding time stamp to the control module.

[0040] The control module is configured to receive and respond to the tag ID and the corresponding time stamp, control the camera to capture an image frame of the target fish body, perform fish body detection based on the image frame of the target fish body through a preset fish body target detection model, set a visual ID based on the detected fish body information, bind the tag ID and the visual ID, and control the camera to track and monitor the target fish body through the visual ID and a preset multi-target tracking algorithm.

[0041] In the embodiment of the present application, the background plate can be a material with a low reflectivity, thereby helping to improve the contrast between the fish image and the background and improve the accuracy of subsequent image analysis. The specific color can be selected according to the actual situation. In actual use, a plurality of RFID card reading coils can be arranged around the area of the side wall, thereby improving the reading efficiency of the tag ID carried by the fish. The shooting range of the camera includes the area where the background plate and the RFID card reading coil are located, thereby facilitating the camera to collect the image of the fish read by the RFID card reading coil. The RFID card reading coil and the camera and the control module are in communication connection, thereby facilitating the control module to receive the tag ID read by the RFID card reading coil, and the camera and the control module to receive the image collected by the camera. The camera can include one or more, and the shooting range of the one or more cameras can cover the entire device corresponding to the three-dimensional structure area, thereby being able to synchronously collect the motion track of the fish in the area and the positional relationship with the RFID card reading coil. The device includes an open inlet and outlet, thereby facilitating the fish to enter and exit. In use, the device can be half-submerged in water, specifically, the camera is above the water surface, and the RFID card reading coil can be located below the water surface. In one example, in order to simulate a natural breeding environment, the monitoring water tank or the specific breeding area in the present application is an open structure, allowing fish to freely enter and exit. The size of the water tank can be adjusted according to the breeding scale and the size of the fish. The inner wall of the water tank should be made of low-reflective and non-reflective materials to reduce optical interference. At the same time, the material of the inner wall of the water tank should also be selected according to the color characteristics of the type of fish being raised. In the side wall or the bottom of the water tank, a plurality of width-appropriate channel-type openings are provided as paths for the fish to enter and exit the monitoring area, for example, the specifications of the openings can be 20-30 cm. The design of these openings should ensure that the fish can pass smoothly, while providing space for the arrangement of RFID coils. Uniform non-direct light sources should be provided above the water tank to avoid strong reflection or shadows on the water surface, thereby optimizing the recognition effect of machine vision. LED array lights can be used, supplemented by soft light covers.

[0042] The RFID card reading coil detects the tag ID carried by the target fish body, and can realize non-contact two-way data communication through wireless radio frequency identification technology through wireless radio frequency. The wireless radio frequency is used to read and write the recording medium (electronic tag or radio frequency card), so as to achieve the purpose of identification and data exchange. After the RFID card reading coil detects the tag ID carried by the target fish body, the tag ID can be sent to the control module for subsequent processing. Specifically, the RFID module in the present application is used to identify the tag carried by the fish body when it passes through a specific area, and obtain its unique identity ID. A plurality of RFID card reading coils are uniformly embedded on the inner wall of each entrance and exit channel. The specific number and interval are determined according to the channel width and the effective identification distance of the card reader, so as to ensure that the tag of the fish body can be stably read by at least one coil when it passes through. The coil should be waterproof packaged. The RFID card reading coil of each channel is connected to one or more multi-channel RFID readers. The reader is connected with the control module through wired or wireless mode, such as Ethernet, Wi-Fi and the like, and transmits the RFID tag ID and reading time stamp in real time. Among them, the fish body wears a passive RFID tag, which is small in size, light in weight and has no obvious influence on the activity of the fish body. The tag can be implanted subcutaneously (for example, under the dorsal fin muscle) or attached externally on the back of the fish body (for example, fixed by biological glue) through biocompatible materials (such as medical silica gel). A unique digital ID is preset in the tag. In one example, a plastic RFID tag with a size of 15mm x 15mm can be selected.

[0043] The control module can receive the tag ID of the fish body read by the RFID card reading coil through the communication connection between the control module and the RFID card reading coil, and control the camera in response to the tag ID. When the control module controls the camera, an image acquisition instruction can be sent to the camera, so that one or more images are acquired through the camera. Since the image acquisition area of the camera includes the area where the RFID card reading coil is located, the image acquired through the camera can include the image of the fish body corresponding to the tag ID read by the RFID card reading coil. In actual use, one or more images can be acquired through the camera, and the acquired images can be preprocessed and screened, such as selecting the image with the highest clarity for further fish body identification. In one example, one or more high-resolution (for example, 4K or higher), high-frame-rate (for example, 30fps or higher) industrial cameras can be erected above the water tank. The field of view of the camera should be able to completely cover the entire monitoring water tank / area, ensuring no dead angles. In actual use, multiple cameras can be used. Multiple cameras can be used to eliminate occlusion and provide more accurate 3D positioning information. The camera should have good low-light performance and waterproof and anti-fog functions. The camera is connected to the image acquisition card through a high-speed data line (such as USB3.0 (a USB interface specification) or GigE Vision (a camera interface standard developed based on Gigabit Ethernet communication protocol)), and the image acquisition card transmits the video stream to the control module.

[0044] In the process of setting a visual ID for the target fish body according to the image frame of the target fish body, the fish body can be identified through a preset fish body position recognition model. Specifically, the fish body recognition model can be implemented as a CNN (Convolutional Neural Network) model, a YOLO (You Only Look Once) model, a PRAM (Position Recognition Arbitrary Location Model), a SpatialLM (an open source 3D visual large model), etc. Thus, a corresponding visual ID is set for the identified model. The label ID and the visual ID are bound. Then, the visual ID is further used to control the camera to track and monitor the target fish body. Thus, in the subsequent tracking and detection process of the target fish body, the detected information can be matched with the visual ID, and the visual ID can be matched with the corresponding label ID. Compared with the scheme that can only identify the fish body through the label ID and cannot continuously track the fish body, the method of the present application can realize the identification and continuous tracking of the fish body. Specifically, when the fish body carrying the RFID tag enters the effective identification range of the RFID card reading coil, the RFID reader immediately reads the label ID and sends a reading signal and a timestamp to the control module. At the same time, after receiving the RFID signal, the control module synchronously instructs the top camera to collect high-definition image frame sequences at the current time and within a few seconds before and after (for example, 2 seconds before and after, a total of 4 seconds). Real-time image processing is performed on the synchronously collected image frames. First, a fish body target detection model based on deep learning (for example, YOLOv8, Faster R-CNN, etc.) is used to accurately detect all fish bodies in the image, and the position and size of each fish body are framed. For each detected fish body, the system generates a unique "visual ID" (for example, a serial number). This visual ID is assigned based on the appearance features (such as color, texture, shape, etc., if there are slight differences) of the fish body or its initial position in the image frame. The successfully associated visual ID and RFID ID are stored in the database to establish a unique identity mapping relationship.

[0045] By the visual ID and the preset multi-target tracking algorithm, when the camera controls tracking and monitoring of the target fish body, once the fish body enters the monitoring area and completes the initial identity binding, subsequent tracking mainly relies on machine vision, and no longer frequently relies on RFID reading, so as to reduce interference and improve efficiency. The multi-target tracking algorithm (for example, DeepSORT, ByteTrack, etc.) is used to track the motion trajectory of the fish body entering the area. These algorithms can maintain the continuity of the visual ID through inter-frame feature matching and motion prediction, even in the case of temporary occlusion or mutual intersection of fish bodies, and record the real-time position coordinates of each fish. Based on the continuously tracked fish body trajectory and position information, the system can further analyze: (1) active range and time: the active area of each fish in the tank, the residence time, etc. (2) the density of the fish school, the distance between individuals, the clustering mode, etc. (3) By identifying the opening and closing action of the fish mouth, the residence time of the fish body in the feed delivery area, and the specific posture of the head near the water surface or feed particles, etc., the occurrence of feeding behavior is judged. This requires more detailed posture estimation or behavior recognition models. (4) Combined with the feed delivery amount and the feeding behavior frequency / duration of the fish body, the individual feeding amount is estimated. This may require additional sensors (such as water surface feed amount monitoring) or more complex behavior models.

[0046] As can be seen, through the device of the embodiments of the present application, only the background plate, the RFID card reading coil and the camera are needed, and after the RFID card reading coil reads the tag ID of the fish body, the camera sets a visual ID for the fish body, thereby binding the tag ID and the visual ID, realizing continuous tracking and monitoring of the fish body through the visual ID, and solving the problems of low efficiency and high cost in the fish body positioning and tracking process.

[0047] In a possible implementation, the control module is specifically configured to, when the camera captures an image frame containing multiple fish bodies, calculate distances between each fish body and the RFID card reading coil through a preset fish body position recognition model; according to the calculated distances, take the fish body closest to the RFID card reading coil as a target fish body, and set the visual ID for the target fish body.

[0048] Since the RFID device is usually installed in the dorsal or ventral muscle of the fish, it can ensure that the tag will not be affected by the movement of the fish body and reduce the impact on the fish. Therefore, in the embodiments of the present application, when calculating the distance between each fish body and the RFID reading coil, the distance between the position where the tag is installed in the fish body and the RFID reading coil can be calculated. Alternatively, the distance between the specified position of the fish body and the RFID reading coil can also be calculated, such as the center position of the fish body, the head of the fish, the tail of the fish, etc. When calculating the distance between each fish body and the RFID reading coil, only the distance between the fish body and the RFID reading coil in the collected picture can be identified, or the spatial coordinates can be identified, so that the calculation is performed according to the identified spatial coordinates. Specifically, the calculation can be performed by a preset fish body position identification model. In one example, the position of the RFID reading coil can be a known position when calculating. In actual use, when there are multiple RFID reading coils, the distance between each fish body and the RFID reading coil can be calculated for each RFID reading coil. Since when there are multiple fish bodies around the RFID reading coil, the tag ID corresponding to the first identified fish body is usually sent to the control module, and the first identified fish body is usually the fish body closest to the RFID reading coil. Therefore, the fish body closest to the RFID reading coil is taken as the target fish body, and the visual ID is set for the target fish body, so that the tag ID corresponding to the fish body closest to the RFID reading coil and the visual ID are bound, thereby facilitating subsequent continuous observation. In actual use, if it is not possible to distinguish which fish body is closest to the RFID reading coil, the fish body can be set with a visual ID and then randomly associated. Specifically, the system can calculate the distance from the geometric center of each fish body to the center point of the RFID reading coil according to the pixel coordinates of the fish body in the image and the physical coordinates of the RFID reading coil calibrated in advance (mapped to the image pixel coordinate system through geometric transformation). In the synchronously collected image frame, if multiple fish bodies exist in a single RFID reading area at the same time, the system will preferentially select the fish body closest to the center of the RFID reader as the association object of the currently read RFID tag ID. If there are multiple fish bodies with a distance very close to the center of the RFID reader (for example, within a set threshold range, such as 5 pixel points), it is difficult to distinguish by position, then the system will temporarily mark the visual ID and the RFID ID of these multiple fish bodies as "to be associated". The system will continue to track these "to be associated" fish bodies, and combine their subsequent behavior patterns (such as which fish leaves the area) or clearer picture information to correct or randomly assign.

[0049] In a possible implementation, the control module is further configured to, when the target fish body is detected by the camera and leaves the shooting range of the camera, unbind the tag ID and the visual ID. In this application, the shooting range of the camera can be larger than the collection range of the RFID card reading coil, so even if the fish body leaves the collection range of the RFID card reading coil, it can continue to be observed by the visual ID. The fish body is continuously tracked based on the initial visual ID by machine vision only, and no longer relies on RFID reading. When the fish body leaves the identification range of the camera, the binding between the tag ID and the visual ID can be unbound because detection cannot continue, so that the identity information of the fish body can be reliably obtained when the fish body enters the monitoring area again next time. Specifically, when the fish body leaves the monitoring area, the identity binding is unbound, and is ready for next time. When the fish body passes through the exit channel again, the RFID tag of the fish body is identified by the card reader again. The control module searches for the associated visual ID according to the RFID reading signal again. Once it is confirmed that the fish body has left the monitoring area, the system unbinds the visual ID and the RFID ID. The visual ID state of the fish body is marked as “not in the monitoring area” so as to re-identify and bind the identity when entering next time.

[0050] In a possible implementation, the control module is specifically configured to collect images of the target fish body at multiple time points by the camera; and identify behavior data of the target fish body according to the collected images at the multiple time points, wherein the behavior data includes a visual ID, a position, and a behavior trajectory. After the images of the target fish body at the multiple time points are collected by the camera, the position and shape of the fish body in each image can be identified, and then the images are arranged in the order of time corresponding to each image, so that the behavior trajectory of the target fish body can be obtained. The current position can be the position corresponding to each time point. When the position of the fish body is calculated, a fish body position recognition model obtained by pre-training can be used for calculation, such as a PRAM (Position Recognition Arbitrary Location Model), a SpatialLM (an open source 3D visual large model), and the like. In actual use, the behavior data of the fish body can be collected to accurately analyze the feeding behavior and feeding amount information of each fish in the fish school, so as to have a reference effect on optimizing the feed formula, the feeding strategy, improving the feed utilization rate, and cultivating excellent strains that are resistant to coarse feed and save feed through genetic improvement.

[0051] In a possible implementation, the device is a frame structure; the background plate is arranged at the bottom of the frame structure; the RFID reading coil is arranged at the sidewall of the frame structure; the camera is arranged at the top of the frame structure; and the frame structure comprises an open entrance and an open exit. The fish body positioning and tracking device in the application can refer to Figure 1 , Figure 1 Another structural diagram of the fish body positioning and tracking device provided by the embodiment of the application is shown in FIG. 3. The structure can be a frame structure, allowing the fish to freely enter and exit, and one or more surfaces can be arranged with RFID coils. In actual use, the device can be immersed in water, thereby facilitating the entry and exit of the fish body.

[0052] It can be seen that, by using the device corresponding to the embodiment of the application, the fish body activity freedom can be realized without being limited in an open feeding environment; the behavior data of a single fish body can be accurately collected in a complex environment such as water optical interference and fish body shielding; the advantages of RFID and machine vision are combined, the continuity of tracking and the accuracy of individual identification are improved, the system complexity and hardware cost are reduced, and the device has a good application and promotion prospect.

[0053] In order to illustrate the scheme of the embodiment of the application, the following describes a specific embodiment, and the specific use method of the device comprises the following steps.

[0054] (1) When the fish body enters or exits the RFID coil, the RFID system reads the tag ID carried by the fish body, and synchronously collects an image frame by the top camera;

[0055] (2) An image analysis algorithm assigns a visual ID to the fish body appearing in the picture based on the picture;

[0056] (3) The visual ID and the RFID are associated by analyzing the positional relationship between the fish body in the image and the RFID coil;

[0057] (4) If there are multiple fish bodies appearing simultaneously in a single RFID reader area, the fish body closest to the center of the RFID reader is preferentially selected as the association object; if it is not possible to distinguish, the fish bodies are randomly associated;

[0058] (5) After the fish body enters the monitoring area, subsequent continuous tracking is performed based on the initial visual ID by machine vision, and the RFID reading is no longer relied on.

[0059] (6) When the fish body passes through the RFID coil and exits the monitoring area, the visual ID and the RFID are disassociated, so that the identity information of the fish body can be reliably obtained when the fish body enters the monitoring area again next time.

[0060] The scheme of the embodiment has the following advantages: (1) combining the identity certainty of RFID and the real-time tracking capability of machine vision, the recognition and tracking accuracy of individual fish bodies in a complex water environment is significantly improved. (2) The traditional physical isolation interference to the natural behavior of fish bodies is avoided, and the characteristics such as aggregation and food competition of the fish school are more truly reflected. (3) Compared with the traditional method which needs a large amount of manual operation or expensive special equipment, the system can realize automatic and continuous data acquisition after deployment, and reduce the long-term operation cost. (4) Not only the individual identity can be recognized, but also various data such as motion trajectory, activity range and feeding behavior frequency can be accurately collected, which provides a scientific basis for feed formula optimization, accurate feeding and breeding strategy adjustment. (5) The long-term tracking data can be used to evaluate key indicators such as feed utilization rate and growth speed of different strains of fish, and support genetic improvement and selection of excellent strains such as coarse material tolerance and feed saving.

[0061] The second aspect of the embodiment of the application provides a fish body positioning and tracking method, referring to Figure 2 , Figure 2 A flowchart of the fish body positioning and tracking method corresponding to the embodiment of the application is applied to a control module in a fish body positioning and tracking device, and the method comprises the following steps:

[0062] In step S21, the image frame of the target fish body is collected by the camera in response to the tag ID and the corresponding timestamp.

[0063] In step S22, the fish body is detected according to the image frame of the target fish body by using a preset fish body target detection model.

[0064] In step S23, the visual ID is set according to the detected fish body information.

[0065] In step S24, the tag ID and the visual ID are bound.

[0066] In step S25, the target fish body is tracked and monitored by the camera by using the visual ID and a preset multi-target tracking algorithm.

[0067] Specifically, in the actual use process, the method can be applied to the control module in the fish body positioning and tracking device, and is implemented by the control module. The control module can be a computer, a server or the like, and the tag ID of the target fish body collected by the RFID card reading coil and the picture collected by the camera can be received by the control module, and the camera can be controlled by the control module.

[0068] In a possible implementation, the visual ID is set for the target fish body according to the image frame of the target fish body, and the method comprises the following steps:

[0069] When the camera captures an image frame containing multiple fish bodies, the distance between each fish body and the RFID card reading coil is calculated by a preset fish body position recognition model.

[0070] According to the calculated distance, the fish body closest to the RFID card reading coil is taken as a target fish body, and the visual ID is set for the target fish body.

[0071] In a possible implementation, the method further includes:

[0072] When the target fish body is detected by the camera and leaves the shooting range of the camera, the binding of the tag ID and the visual ID is released.

[0073] In a possible implementation, the control module is specifically configured to: acquire images of the target fish body at multiple time points by the camera; and identify behavior data of the target fish body according to the acquired images at the multiple time points, wherein the behavior data includes a visual ID, a position, and a behavior track.

[0074] It can be seen that, by the method of the embodiments of the present application, after the tag ID of the fish body is read by the RFID card reading coil, the visual ID of the fish body is set by the camera, so that the tag ID and the visual ID are bound, and the fish body is continuously tracked and monitored by the visual ID, thereby solving the problems of low efficiency and high cost in the fish body positioning and tracking process.

[0075] The embodiments of the present application also provide an electronic device, and the device includes:

[0076] a memory for storing a computer program;

[0077] a processor for executing the program stored on the memory to implement the following steps:

[0078] receiving and responding to the tag ID and the corresponding timestamp to control the camera to acquire an image frame of the target fish body;

[0079] detecting the fish body according to the image frame of the target fish body by a preset fish body target detection model;

[0080] setting a visual ID according to the detected fish body information;

[0081] binding the tag ID and the visual ID;

[0082] controlling the camera to track and monitor the target fish body by the visual ID and a preset multi-target tracking algorithm.

[0083] The communication bus mentioned in the above electronic device can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0084] The communication interface is used for communication between the above electronic device and other devices.

[0085] The memory can include a Random Access Memory (RAM) and can also include a Non-Volatile Memory (NVM), such as at least one disk memory. Optionally, the memory can also be at least one storage device located away from the aforementioned processor.

[0086] The processor mentioned above can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0087] In another embodiment provided in the present application, a computer readable storage medium is also provided, and the computer readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of any fish body positioning and tracking method described above are implemented.

[0088] In another embodiment provided in the present application, a computer program product containing instructions is also provided, and when the computer program product is run on a computer, the computer is caused to execute any fish body positioning and tracking method in the above embodiments.

[0089] In the embodiments described above, all or some of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or some of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded into and executed by a computer, all or some of the processes or functions according to the embodiments described in the specification are generated. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through wired (for example, coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (for example, infrared, wireless, microwave, etc.) manner. The computer readable storage medium can be any available medium accessible by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be magnetic media (for example, floppy disk, hard disk, magnetic tape), optical media (for example, DVD), or solid state disk (SSD) and the like.

[0090] It should be noted that, in this document, the terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "including a" does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0091] Each of the embodiments in the specification is described in a related manner, and the same or similar parts between each of the embodiments can be referred to each other. Each of the embodiments focuses on the difference from other embodiments. In particular, for the method, electronic device and storage medium embodiments, since they are basically similar to the device embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiment.

[0092] The above merely provides the preferred embodiment of the present application, and not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A fish body positioning tracking device, characterized by, The device comprises: a background plate arranged at the bottom, an RFID card reading coil arranged at the side wall, a camera arranged at the top, the shooting range of the camera including the area where the background plate and the RFID card reading coil are located, the RFID card reading coil and the camera are in communication connection with the control module, and the device includes an open entrance and exit; the RFID card reading coil is used to detect the label ID carried by the target fish body and send the label ID and the corresponding time stamp to the control module; the control module is used to receive and respond to the label ID and the corresponding time stamp, control the camera to collect the image frame of the target fish body, perform fish body detection according to the image frame of the target fish body through a preset fish body target detection model, set a visual ID according to the detected fish body information, bind the label ID and the visual ID, and control the camera to track and monitor the target fish body through the visual ID and a preset multi-target tracking algorithm; the control module is specifically used to collect images of the target fish body at multiple time points through the camera, identify behavior data of the target fish body according to the collected images at multiple time points, wherein the behavior data includes a visual ID, a position, and a behavior trajectory, calculate the distance between each fish body and the RFID card reading coil through a preset fish body position recognition model when the camera collects an image frame containing multiple fish bodies, set the visual ID for the fish body closest to the RFID card reading coil as the target fish body according to the calculated distance, and unbind the label ID and the visual ID when the camera detects that the target fish body leaves the shooting range of the camera.

2. The apparatus of claim 1, wherein, The device is a frame structure; the background plate is arranged at the bottom of the frame structure, the RFID card reading coil is arranged at the side wall of the frame structure, and the camera is arranged at the top of the frame structure.

3. A fish body positioning tracking method characterized by, The method applied to the control module of the fish body positioning and tracking device of claim 1 comprises: receiving and responding to the label ID and the corresponding time stamp, controlling the camera to collect the image frame of the target fish body; performing fish body detection according to the image frame of the target fish body through a preset fish body target detection model; setting a visual ID according to the detected fish body information; binding the label ID and the visual ID; controlling the camera to track and monitor the target fish body through the visual ID and a preset multi-target tracking algorithm; when the camera detects that the target fish body leaves the shooting range of the camera, unbinding the label ID and the visual ID. The fish body target detection model is used to detect the fish body according to the image frame of the target fish body, including: the camera is used to collect images of the target fish body at multiple time points; behavior data of the target fish body is identified according to the collected images at multiple time points, wherein the behavior data includes visual ID, position, and behavior trajectory. The visual ID is set for the target fish body according to the image frame of the target fish body, including: when the camera collects an image frame containing multiple fish bodies, a preset fish body position identification model is used to calculate the distance between each fish body and the RFID card reading coil; according to the calculated distance, the fish body closest to the RFID card reading coil is taken as the target fish body, and the visual ID is set for the target fish body.

4. An electronic device, comprising: Including: A memory for storing a computer program; A processor for executing the program stored on the memory to realize the method of claim 3.

5. A computer readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the method of claim 3.

Images