Animal weight prediction method, device and equipment and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, and discloses an animal weight prediction method and device, equipment and a storage medium, and the method comprises the steps: obtaining a pre-collected animal image; acquiring image parameters of the animal image, and determining whether the animal image meets a preset shooting condition according to the image parameters; and if the animal image meets the shooting condition, inputting the animal image and the image parameters into a pre-trained weight prediction model, and predicting the weight of the target animal according to the animal image and the image parameters by adopting the weight prediction model. The method is suitable for the field of financial services, can measure the weight of the animal more accurately, and improves the efficiency and accuracy of loss assessment of insurance claim settlement.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology and is applied to the scenario of financial technology business processing, and particularly relates to a method, device, equipment and storage medium for predicting animal weight. Background Art

[0002] In the claim settlement and loss assessment link in the insurance field, especially for livestock insurance such as pig farming, cattle, horses and other animal insurances, how to accurately and efficiently evaluate the damage situation of animals is crucial for insurance companies and farmers. Traditionally, in the process of claim settlement and loss assessment, for the measurement of animal weight, it relies on manual operation of traditional weighing scales. This method requires a large amount of manpower and material resources to move and weigh each animal, and it is prone to human errors during the operation process, resulting in low accuracy and low efficiency.

[0003] In recent years, with the development of artificial intelligence (AI) technology, the industry has begun to try to apply AI weighing technology to replace traditional weighing scale weighing. However, the existing AI weighing technology has deficiencies in image acquisition and parameter processing, which in turn affects the accurate calculation of animal weight by the AI algorithm, resulting in adverse effects.

[0004] In view of this, it is necessary to provide an improved solution for animal weighing to solve the problem of how to more accurately measure the weight of animals, thereby improving the efficiency and accuracy of claim settlement and loss assessment. Summary of the Invention

[0005] The present invention provides a method, device, computer equipment and storage medium for predicting animal weight, aiming to solve the technical problem of how to more accurately measure the weight of animals, and improving the efficiency and accuracy of insurance claim settlement and loss assessment.

[0006] In the first aspect, a method for predicting animal weight is provided, including:

[0007] Obtaining a pre-collected animal image;

[0008] Obtaining the image parameters of the animal image, and determining whether the animal image meets a preset shooting condition according to the image parameters;

[0009] If the animal image meets the shooting condition, inputting the animal image and the image parameters into a pre-trained weight prediction model, and using the weight prediction model to predict the target animal weight according to the animal image and the image parameters.

[0010] In the second aspect, a device for predicting animal weight is provided, including:

[0011] An image acquisition module, configured to obtain a pre-collected animal image;

[0012] A parameter acquisition module, configured to acquire image parameters of the animal image, and determine whether the animal image meets a preset shooting condition according to the image parameters;

[0013] A weight prediction module, configured to, if the animal image meets the shooting condition, input the animal image and the image parameters into a pre-trained weight prediction model, and use the weight prediction model to predict the weight of the target animal according to the animal image and the image parameters.

[0014] In a third aspect, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above animal weight prediction method are implemented.

[0015] In a fourth aspect, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above animal weight prediction method are implemented.

[0016] In the solutions implemented by the above animal weight prediction method, device, computer device, and storage medium, an animal image can be obtained by performing a standardized shooting on a target animal. By determining whether the animal image meets a preset shooting condition according to the image parameters of the animal image, the animal images can be screened according to the preset shooting condition, ensuring that only the images that meet the shooting condition are used for subsequent weight prediction, thereby improving the prediction accuracy; by inputting the animal image that meets the shooting condition and the image parameters into a pre-trained weight prediction model, the weight prediction model can quickly and accurately predict the weight of the target animal according to the animal image and the image parameters. Based on the solution of this application, the accuracy of processing animal images and image parameters is effectively improved, the loss caused by weighing errors is reduced, and further the efficiency and accuracy of insurance claim settlement are improved. Description of the Drawings

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 is an exemplary system architecture diagram of an animal weight prediction method in an embodiment of the present invention;

[0019] Figure 2 is a flowchart of an animal weight prediction method in an embodiment of the present invention;

[0020] Figure 3 isFigure 1 Schematic flowchart of a specific implementation manner of step S60 in

[0021] Figure 4 Schematic structural diagram of an animal weight prediction device in an embodiment of the present invention;

[0022] Figure 5 Schematic structural diagram of a computer device in an embodiment of the present invention. Specific implementation manner

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs; the terms used in the description of the present application in this specification are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the description and claims of this application and the above drawings are intended to cover non-exclusive inclusion. The terms "first", "second", etc. in the description and claims of this application or the above drawings are used to distinguish different objects and not to describe a specific order.

[0024] Referring to "embodiments" herein means that specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of this application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0025] In order to enable those skilled in the technical field to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the drawings.

[0026] As Figure 1 shown, the system architecture 100 may include a terminal device 101, a network 102, and a server 103. The terminal device 101 may be a laptop computer 1011, a tablet computer 1012, or a mobile phone 1013. The network 102 is a medium for providing a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0027] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications may be installed on the terminal device 101, such as a web browser application, a shopping application, a search application, an instant messaging tool, an email client, a social platform software, etc.

[0028] The terminal device 101 can be various electronic devices with a display screen and supporting web browsing. In addition to the laptop 1011, tablet computer 1012, or mobile phone 1013, the terminal device 101 can also be an e-book reader, MP3 player (Moving Picture Experts Group Audio Layer III), MP4 (Moving Picture Experts Group Audio Layer IV) player, laptop portable computer, desktop computer, and so on.

[0029] The server 103 can be a server that provides various services, such as a background server that supports the pages displayed on the terminal device 101.

[0030] It should be noted that the animal weight prediction method provided by the embodiments of the present application is generally executed by the server / terminal device. Correspondingly, the animal weight prediction device is generally set in the server / terminal device.

[0031] It should be understood that Figure 1 the numbers of the terminal devices, network, and server in

[0032] Continue to refer to Figure 2 as shown in Figure 2 which is a schematic flowchart of the animal weight prediction method provided by the embodiments of the present invention, including the following steps:

[0033] S20: Obtain the pre-collected animal images;

[0034] The animal weight prediction method provided by the present invention can be applied to the intelligent animal weight prediction engine in various application scenarios. The intelligent animal weight prediction engine internally integrates a weight prediction model. The intelligent animal weight prediction engine is usually implemented by a server, which is connected to terminal devices such as mobile phone devices or other camera devices through a network, and can obtain animal images collected based on mobile phone devices or other camera devices in real time. Among them, the mobile phone device or other camera devices include a camera, a gyroscope, and a display screen. That is, in the mobile phone device or other camera devices, the camera, the gyroscope, and the display screen are key hardware components in the device, jointly supporting their functions in aspects such as taking pictures, positioning, and displaying. The camera is a component on the mobile phone device or other camera devices used to capture and record images and videos. The gyroscope is a sensor used to measure the angular velocity and direction of the mobile phone device or other camera devices in space. The display screen is a touch screen display interface on the mobile phone device or other camera devices used to display information such as images, texts, and videos.

[0035] Specifically, taking a mobile device as an example, the process of collecting animal images using a mobile device can be as follows: An application for animal weight prediction is pre-installed on the mobile device. This application can call the mobile camera to take pictures, read the data of the mobile gyroscope, and display an auxiliary positioning tool - an animal contour map and a reference checkerboard on the mobile device's display screen. After opening this application, a preset animal contour map and a reference checkerboard are displayed on the mobile device's display screen. Use the mobile camera to aim at the animal and take pictures. During the shooting process, the application guides the user to adjust the shooting angle in real time by reading the data of the mobile gyroscope to ensure that the shooting angle meets the preset requirements (such as vertical angle, horizontal angle, etc.) and reduce image distortion caused by angle problems. At the same time, during the shooting process, through the animal contour map on the mobile device's display screen, guide the user to place the animal within the set range of the contour map; through the reference checkerboard on the mobile device's display screen, guide the user to compare the size of the reference checkerboard with the actual checkerboard placed directly below the animal to ensure minimizing image deformation. After completing the shooting of the animal image, the application transmits the captured animal image to the intelligent animal weight prediction engine (i.e., the server), and the server obtains the animal image collected based on the mobile device.

[0036] It should be noted that before obtaining the pre-collected animal image, it may also include the process of taking pictures of the animal using a mobile device or other camera devices. Specifically, before step S20, that is, before obtaining the pre-collected animal image, the following steps may also be included:

[0037] In response to the start signal triggered by the user, control the camera and gyroscope to turn on and display the shooting page, and the shooting page is provided with an animal contour map, a reference checkerboard, a shooting angle display area, a focal length adjustment slider, and a wide-angle switch;

[0038] Specifically, when a user operates a mobile device or other camera device to open a pre-installed application, the device sends a startup signal to the intelligent animal weight prediction engine (i.e., the server). At this time, in response to the startup signal, the server generates corresponding control commands to control the camera and gyroscope on the mobile device or other camera device to turn on and display the shooting page. Among them, the shooting page is set with an animal contour diagram, a reference checkerboard, a shooting angle display area, a focal length adjustment slider, and a wide-angle switch. The animal contour diagram is used to guide the user to correctly place the animal within the shooting range of the contour diagram, ensuring that the position and posture of the animal in the image meet the requirements of subsequent analysis; the reference checkerboard is used to compare with the actual checkerboard placed directly below the animal during the shooting process to avoid distortion of the size and shape of the image; the shooting angle display area is used to display in real time the shooting angle collected by the gyroscope; the focal length adjustment slider is used to adjust the focal length multiple of the camera; the wide-angle switch is used to turn on or off the wide angle of the camera.

[0039] Optionally, after starting the camera and displaying the shooting page on the display screen, a pre-set focal length adjustment slider is displayed on the shooting page. The focal length adjustment slider can be a horizontal or vertical slider bar to dynamically adjust the focal length multiple of the mobile phone camera when adjusting the position of the slider bar. Use image processing techniques (such as edge detection, shape matching, etc.) to detect the target animal contour in the shooting page, and compare the detected target animal contour with the pre-set animal contour diagram. When the target animal contour exceeds the pre-set animal contour diagram, slide the focal length adjustment slider to adjust the focal length multiple so that the target animal contour in the shooting page is within the range of the pre-set animal contour diagram for shooting. As the focal length multiple changes, the size of the animal contour in the shooting image will be adjusted accordingly until the animal contour returns to the range of the pre-set animal contour diagram.

[0040] Optionally, the focal length adjustment slider displayed on the shooting page can also support manual sliding operations by the user. When the user slides the focal length adjustment slider, the focal length multiple of the mobile phone camera is adjusted accordingly.

[0041] By sliding the focal length adjustment slider when the target animal contour in the shooting page exceeds the pre-set animal contour diagram, the focal length multiple is adjusted in real time, so that the animal contour can be accurately matched within the animal contour diagram. This real-time adjustment function can ensure that the captured animal image always meets the range of the pre-set animal contour diagram. In addition, through focal length adjustment, the animal contour can be quickly adjusted to the appropriate range, avoiding the process of multiple shootings and post-processing, thereby improving the shooting efficiency.

[0042] Optionally, after starting the camera and displaying the shooting page on the display screen, a pre-set wide-angle switch is displayed on the shooting page. The application captures images in real time through the camera and analyzes the size of the animal in the image using image processing techniques (such as edge detection, shape matching, etc.). For example, through edge detection, the edge information of the target animal in the image is identified to determine the contour of the target animal; based on the identified edge information, the size of the target animal (such as width, height, and area, etc.) is calculated. The pre-set size threshold is obtained, and the size of the target animal is compared with this size threshold. When the size of the target animal is greater than the size threshold, it indicates that the size of the target animal is too large. At this time, the wide-angle switch is turned on, and the camera switches to the wide-angle mode, that is, a wide-angle lens is used to capture the animal image, ensuring that the animal can completely appear in the shooting page.

[0043] Optionally, the wide-angle switch displayed on the shooting page can also support manual click operations by the user. When the user manually clicks the wide-angle switch, the wide-angle mode of the camera is correspondingly turned on or off.

[0044] By turning on the wide-angle switch when it is recognized that the animal size is too large, it can ensure that the entire animal can be completely included in the shooting picture. This is particularly important for application scenarios such as insurance claim assessment that require complete recording of animal conditions. By using wide-angle shooting, users can shoot large-size objects at a relatively close distance without having to step back to obtain a complete picture. This helps in shooting in a limited space, especially in environments such as pigsties or farms, significantly enhancing the flexibility and adaptability of shooting and improving the user experience.

[0045] In response to the shooting signal triggered by the user, the animal image captured by the camera, the target shooting angle, focal length multiple, and wide-angle state collected by the gyroscope are obtained, where the animal image includes the animal contour map and / or the reference checkerboard.

[0046] Specifically, when the user triggers the shooting area on the mobile device or other camera device, a shooting signal is generated. In response to the shooting signal through the application, the camera is started to capture images, and the animal image captured by the camera is obtained, where the animal image includes the animal contour map and / or the reference checkerboard; at the same time, the application starts the gyroscope for angle monitoring to obtain the target shooting angle monitored by the mobile phone gyroscope, including vertical angle, horizontal angle, etc.; the application can also obtain the current focal length multiple and wide-angle state. The obtained animal image, target shooting angle, focal length multiple, and wide-angle state are transmitted to the intelligent animal weight prediction engine (i.e., the server) through the application, and the server obtains image parameters such as the animal image, target shooting angle, focal length multiple, and wide-angle state.

[0047] By presetting an animal contour map on the shooting page, the correct placement position and posture of the animal during shooting can be more intuitively guided, which helps to improve the accuracy and consistency of shooting. By presetting a reference checkerboard, it is convenient for subsequent angle correction and image matching, and improves the accuracy of image processing. By responding to the shooting signal triggered by the user, the animal image captured by the camera and the data such as the target shooting angle, focal length multiple, and wide-angle state collected by the gyroscope are obtained, realizing the synchronous acquisition of data such as images and angles, and ensuring the integrity and timeliness of the data.

[0048] S40: Obtain the image parameters of the animal image, and determine whether the animal image meets the preset shooting conditions according to the image parameters;

[0049] After the animal image is captured, the mobile device or other imaging device transmits the image parameters corresponding to the captured animal image to the intelligent animal weight prediction engine (i.e., the server). The server obtains the image parameters corresponding to the animal image, and the image parameters may include, but are not limited to, the target shooting angle (such as vertical angle, horizontal angle) collected by the gyroscope, the animal contour map, the target animal contour, the reference checkerboard, the actual checkerboard, the focal length multiple, the wide-angle state, the equivalent focal length value, the camera focal length value, the width of the camera sensor, the height of the camera sensor, and the size of the target animal, etc.

[0050] Obtain the preset shooting conditions. Among them, the shooting conditions may include, but are not limited to, shooting angle conditions, contour coincidence conditions, checkerboard coincidence rate conditions, etc. For example, the shooting angle condition can be set as "the vertical angle is within ±10°, and the horizontal angle is within ±5°", the contour coincidence condition can be set as "the contour coincidence rate between the animal contour map and the animal is not less than 80%", and the checkerboard coincidence rate condition can be set as "the coincidence rate between the reference checkerboard and the actual checkerboard is not less than 90%". Compare the obtained image parameters with the preset shooting conditions one by one. When all the image parameters meet the preset shooting conditions, it is determined that the captured animal image meets the preset shooting conditions; when any one of the image parameters does not meet the preset shooting conditions, it is determined that the animal image does not meet the preset shooting conditions.

[0051] In some embodiments, at least one specific scheme for determining whether the animal image meets the preset shooting conditions is provided. Among them, the preset shooting conditions may include shooting angle conditions. In step S40, obtaining the image parameters of the animal image and determining whether the animal image meets the preset shooting conditions may include the following steps S401 - S403:

[0052] S401: Obtain the target shooting angle of the animal image as the image parameter, and the target shooting angle is collected by the gyroscope;

[0053] S402: Obtain the preset shooting angle condition, and compare the target shooting angle with the shooting angle condition;

[0054] S403: When the target shooting angle meets the shooting angle condition, determine that the animal image meets the shooting condition; otherwise, it does not meet the shooting condition.

[0055] Specifically, obtain the target shooting angle of the animal image collected by the gyroscope as an image parameter. The target shooting angle may include a target vertical angle and a target horizontal angle. Obtain the preset shooting angle condition, which includes a vertical angle range and a horizontal angle range. Compare the target shooting angle with the preset shooting angle condition to determine whether the target vertical angle is within the vertical angle range and whether the target horizontal angle is within the horizontal angle range. When the target shooting angle meets the shooting angle condition, that is, when the target vertical angle is within the vertical angle range and the target horizontal angle is within the horizontal angle range, determine that the animal image meets the shooting condition. Otherwise, when the target shooting angle does not meet the shooting angle condition, that is, when the target vertical angle is not within the vertical angle range or the target horizontal angle is not within the horizontal angle range, determine that the animal image does not meet the shooting condition.

[0056] Optionally, the preset shooting condition may include a contour coincidence condition. In step S40, obtaining the image parameters of the animal image and determining whether the animal image meets the preset shooting condition may further include the following steps S404 - S407:

[0057] S404: Obtain the animal contour map set on the shooting page;

[0058] S405: Extract the target animal contour from the animal image, match the animal contour map with the target animal contour to obtain a contour coincidence area and a contour non - coincidence area;

[0059] S406: Calculate the contour coincidence rate as an image parameter according to the contour coincidence area and the contour non - coincidence area;

[0060] S407: Obtain the preset contour coincidence condition. When the contour coincidence rate meets the contour coincidence condition, determine that the animal image meets the shooting condition; otherwise, it does not meet the shooting condition.

[0061] Specifically, for the captured animal image, obtain the animal contour map set on the capture page. Identify and extract the target animal contour in the animal image by using image processing technology. Match the preset animal contour map with the target animal contour in the animal image to obtain a matching result, which includes a contour overlapping area and a contour non-overlapping area. Among them, the contour overlapping area represents the area where the target animal is placed within the preset animal contour map; the contour non-overlapping area represents the area where the target animal is placed outside the preset animal contour map.

[0062] Specifically, calculate the contour overlapping rate according to the contour overlapping area and the contour non-overlapping area. For example, the contour overlapping rate = contour overlapping area / (contour overlapping area + contour non-overlapping area). Obtain the preset contour overlapping condition, which is a threshold range regarding the contour overlapping rate. Compare the calculated contour overlapping rate with the preset contour overlapping condition to determine whether the contour overlapping rate is within the threshold range set by the contour overlapping condition. When the contour overlapping rate meets the contour overlapping condition, determine that the animal image meets the capture condition. Otherwise, when the contour overlapping rate does not meet the contour overlapping condition, determine that the animal image does not meet the capture condition.

[0063] Optionally, the preset capture condition may include a chessboard overlapping rate condition. The captured animal image includes an actual chessboard and a target animal. In step S40, obtain the image parameters of the animal image, and determine whether the animal image meets the preset capture condition according to the image parameters. It may also include the following steps S408 - S412:

[0064] S408: Obtain the reference chessboard set on the capture page;

[0065] S409: Determine the actual chessboard from the animal image, and the actual chessboard is placed directly below the target animal;

[0066] S410: Match the reference chessboard with the actual chessboard to obtain the target chessboard overlapping rate as the image parameter;

[0067] S411: Obtain the preset chessboard overlapping rate condition, and compare the target chessboard overlapping rate with the chessboard overlapping rate condition;

[0068] S412: When the target chessboard overlapping rate meets the chessboard overlapping rate condition, determine that the animal image meets the capture condition; otherwise, it does not meet the capture condition.

[0069] Specifically, for the captured animal image, obtain the reference checkerboard set on the shooting page. By using image processing technology to identify the actual checkerboard in the animal image, which is placed directly below the target animal. Compare the reference checkerboard with the actual checkerboard in the animal image, and calculate the target checkerboard coincidence rate between the reference checkerboard and the actual checkerboard as the image parameter. The checkerboard coincidence rate can be used to evaluate the degree of image distortion. A high coincidence rate indicates less image distortion, while a low coincidence rate indicates significant distortion or distortion in the image.

[0070] Specifically, obtain the pre-set checkerboard coincidence rate condition, which is the threshold range for the checkerboard coincidence rate. Compare the target checkerboard coincidence rate with the pre-set checkerboard coincidence rate condition to determine whether the target checkerboard coincidence rate is within the threshold range set by the checkerboard coincidence rate condition. When the target checkerboard coincidence rate meets the checkerboard coincidence rate condition, it is determined that the animal image meets the shooting conditions. Otherwise, when the target checkerboard coincidence rate does not meet the checkerboard coincidence rate condition, it is determined that the animal image does not meet the shooting conditions.

[0071] By matching the animal contour map with the target animal contour in the animal image, the contour coincidence area and the contour non-coincidence area are obtained, which can be used to evaluate the accuracy and integrity of the animal image shooting. By matching the reference checkerboard with the actual checkerboard in the animal image, the target checkerboard coincidence rate is obtained, which can be used to evaluate the degree of image tilt and distortion. Based on this embodiment, through refined image acquisition and parameter measurement, the technical problems of poor image acquisition control and insufficient acquisition parameters existing in the animal weighing process are effectively solved. Moreover, by comparing the image parameters with the pre-set shooting conditions, strict control and guarantee of the animal image quality are achieved, the shooting efficiency and accuracy are improved, and a solid foundation is laid for subsequent image processing and applications.

[0072] S60: If the animal image meets the shooting conditions, input the animal image and the image parameters into a pre-trained weight prediction model, and use the weight prediction model to predict the weight of the target animal according to the animal image and the image parameters.

[0073] In this embodiment, when it is determined that the animal image meets the preset shooting conditions, the captured animal image and the image parameters corresponding to the animal image are input into a pre-trained weight prediction model. The weight prediction model is used to predict the body weight based on the animal image and the image parameters to obtain the target animal weight. The weight prediction model is trained based on a large number of historical animal images and historical image parameters, where each historical animal image is labeled with the corresponding animal weight. The image parameters input into the model may include, but are not limited to, one or more of the following data: the target shooting angles (such as vertical angle, horizontal angle) collected by the gyroscope, the contour coincidence rate between the contour coincidence area and the contour non-coincidence area, the chessboard coincidence rate between the reference chessboard and the actual chessboard, the focal length multiple, the wide-angle state, the equivalent focal length value, the camera focal length value, the width of the camera sensor, the height of the camera sensor, and the size of the target animal, etc.

[0074] It should be noted that when the animal image does not meet the preset shooting conditions, or when the acquisition of the image parameters of the animal image is unsuccessful, it is regarded as poor image acquisition control or insufficient parameter acquisition. To ensure that only images that meet the shooting conditions are used for subsequent weight prediction to improve the prediction accuracy, when it is determined that the animal image does not meet the preset shooting conditions, the animal image and the image parameters are re-acquired. Specifically, after step S60, that is, after determining whether the animal image meets the preset shooting conditions according to the image parameters, the following steps S70 - S80 may further be included:

[0075] S70: If the animal image does not meet the shooting conditions, obtain the abnormal parameters that do not meet the shooting conditions;

[0076] S80: Generate a prompt message according to the abnormal parameters to prompt the user to re-acquire the animal image until the re-acquired animal image meets the shooting conditions.

[0077] For steps S70 - S80, when the animal image does not meet the preset shooting conditions, that is, when there are image parameters that do not meet the preset shooting conditions, obtain the image parameters that do not meet the shooting conditions as abnormal parameters. Generate a prompt message according to the obtained abnormal parameters. For example, the prompt message may be "Abnormal shooting angle, please re-acquire the image", "Abnormal contour coincidence rate, please re-acquire the image" or "Abnormal chessboard coincidence rate, please re-acquire the image", etc., to prompt the user to re-acquire the animal image. Until the re-acquired animal image meets the preset shooting conditions, input the re-acquired animal image and the image parameters into the pre-trained weight prediction model, and use the weight prediction model to predict the animal weight based on the re-acquired animal image and the image parameters.

[0078] Through the above solution, not only is the existence of the problem pointed out by the abnormal parameters, but also a basis for subsequent improvement is provided. By identifying and analyzing the abnormal parameters, it is possible to understand which aspects of the shooting conditions are not met, thereby adjusting the shooting strategy or device settings to improve the image quality. By generating a prompt message, the problems existing in the current image can be intuitively prompted, and the user can be guided on how to operate to improve the image quality, which helps to enhance the user's trust and satisfaction with the system. By prompting the user to re-acquire the image, it is possible to ensure that the image data used for subsequent weight prediction meets certain quality requirements, which helps to improve the prediction accuracy and reduce the error caused by poor image quality.

[0079] In some embodiments, referring to Figure 3 as shown, in step S60, using the weight prediction model to predict the weight of the target animal based on the animal image and the image parameters may include the following steps S601 - S603:

[0080] S601: Use the weight prediction model to perform image semantic segmentation on the animal image to obtain a number of target animal regions;

[0081] S602: Obtain the target image parameters corresponding to the target animal regions from the image parameters;

[0082] S603: Calculate the animal body shape characteristics based on the target image parameters, construct a three - dimensional animal portrait based on the animal body shape characteristics, and estimate the weight of the target animal based on the three - dimensional animal portrait.

[0083] For steps S601 - S603, first, obtain a pre - trained weight prediction model. This weight prediction model can identify and distinguish different parts in an animal image, and perform image semantic segmentation on different parts. Among them, different parts can include the torso, head, and limbs of the target animal, etc. Input the captured animal image into the weight prediction model, use the weight prediction model to perform image semantic analysis on the animal image, identify different pixel regions in the animal image, determine different parts of the target animal according to different pixel regions, segment different parts of the target animal to obtain several target animal regions. These target animal regions can include a torso region, a head region, and limb regions, etc. For several target animal regions, obtain the target image parameters corresponding to each target animal region from the image parameters. Use the target image parameters corresponding to each target animal region to calculate the animal body shape characteristics of the target animal. These animal body shape characteristics can include the length, width, height, volume, animal body shape symmetry, and obesity degree of different parts of the target animal, etc. According to the calculated animal body shape characteristics, use 3D modeling technology to construct a virtual 3D animal portrait. Map the characteristics in the constructed 3D animal portrait to a pre - constructed animal weight database, and obtain the closest estimated weight through comparison and matching, which is the weight of the target animal estimated by the model.

[0084] Through image semantic segmentation technology, the model can accurately separate the target animal from a complex background, avoid the interference of background information, and improve the accuracy of subsequent processing. Moreover, precise image semantic segmentation helps the model better understand the shape and structure of the target animal, thus improving the accuracy of weight prediction. By specifically extracting the target image parameters according to the segmented target animal regions, and calculating the animal body shape characteristics based on the target image parameters, a quantitative description of the animal shape is achieved. The constructed 3D animal portrait visually displays the shape and structure of the target animal. Based on the 3D animal portrait and animal body shape characteristics, the model can more accurately estimate the weight of the target animal, effectively improving the accuracy and efficiency of prediction.

[0085] The animal weight prediction method provided by the embodiment of the present invention obtains an animal image by regularly photographing a target animal. Among them, a gyroscope is used to control the shooting angle to ensure that the shooting angle meets the requirements and reduce image distortion caused by angle changes; an animal contour map and a reference checkerboard are used to assist in animal positioning to ensure that the position of the captured animal image is accurate and the contour is clear. By precisely controlling the shooting process, key image parameters can be fully collected; by determining whether the animal image meets the preset shooting conditions according to the image parameters of the animal image, the animal images can be screened according to the preset shooting conditions to ensure that only the images that meet the shooting conditions are used for subsequent weight prediction, thereby improving the accuracy of the prediction; by inputting the animal image and image parameters that meet the shooting conditions into a pre-trained weight prediction model, the weight prediction model can quickly and accurately predict the weight of the target animal according to the animal image and image parameters, effectively improving the accuracy of processing animal images and image parameters, reducing losses caused by weighing errors, and further improving the efficiency and accuracy of insurance claim settlement and loss assessment.

[0086] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0087] In one embodiment, an animal weight prediction device is provided, and the animal weight prediction device corresponds one-to-one with the animal weight prediction method in the above embodiment. As Figure 4 shown, the animal weight prediction device includes an image acquisition module 301, a parameter acquisition module 302, and a weight prediction module 303. The detailed descriptions of each functional module are as follows:

[0088] The image acquisition module 301 is used to acquire a pre-collected animal image;

[0089] The parameter acquisition module 302 is used to acquire the image parameters of the animal image and determine whether the animal image meets the preset shooting conditions according to the image parameters;

[0090] The weight prediction module 303 is used to, if the animal image meets the shooting conditions, input the animal image and the image parameters into a pre-trained weight prediction model, and use the weight prediction model to predict the weight of the target animal according to the animal image and the image parameters.

[0091] In one embodiment, the animal weight prediction device further includes:

[0092] The abnormality acquisition module is used to, if the animal image does not meet the shooting conditions, acquire the abnormal parameters that do not meet the shooting conditions;

[0093] An information prompt module, configured to generate a prompt message according to the abnormal parameter to prompt the user to re - collect an animal image until the re - collected animal image meets the shooting conditions.

[0094] In one embodiment, the shooting conditions include a shooting angle condition. The parameter acquisition module 302 is specifically configured to:

[0095] Obtain the target shooting angle of the animal image as an image parameter, where the target shooting angle is collected by a gyroscope;

[0096] Obtain a preset shooting angle condition, compare the target shooting angle with the shooting angle condition, and when the target shooting angle meets the shooting angle condition, determine that the animal image meets the shooting conditions; otherwise, it does not meet the shooting conditions.

[0097] In one embodiment, the shooting conditions include a contour coincidence condition. The parameter acquisition module 302 is specifically configured to:

[0098] Obtain the animal contour map set on the shooting page;

[0099] Determine the target animal contour from the animal image, match the animal contour map with the target animal contour, and obtain a contour coincidence area and a contour non - coincidence area;

[0100] Calculate a contour coincidence rate as an image parameter according to the contour coincidence area and the contour non - coincidence area;

[0101] Obtain a preset contour coincidence condition, and when the contour coincidence rate meets the contour coincidence condition, determine that the animal image meets the shooting conditions; otherwise, it does not meet the shooting conditions.

[0102] In one embodiment, the shooting conditions include a chessboard coincidence rate condition. The animal image includes an actual chessboard grid and a target animal. The parameter acquisition module 302 is specifically configured to:

[0103] Obtain the reference chessboard grid set on the shooting page;

[0104] Determine the actual chessboard grid from the animal image, and the actual chessboard grid is placed directly below the target animal;

[0105] Match the reference chessboard grid with the actual chessboard grid to obtain a target chessboard coincidence rate as an image parameter;

[0106] Obtain a preset chessboard coincidence rate condition, compare the target chessboard coincidence rate with the chessboard coincidence rate condition, and when the target chessboard coincidence rate meets the chessboard coincidence rate condition, determine that the animal image meets the shooting conditions; otherwise, it does not meet the shooting conditions.

[0107] In one embodiment, the animal weight prediction device further includes:

[0108] A start shooting module, configured to control the camera and gyroscope to turn on and display a shooting page in response to a start signal triggered by a user. The shooting page is provided with an animal contour diagram, a reference checkerboard, a shooting angle display area, a focal length adjustment slider, and a wide-angle switch;

[0109] A trigger shooting module, configured to obtain an animal image captured by the camera, a target shooting angle, a focal length multiple, and a wide-angle state collected by the gyroscope in response to a shooting signal triggered by a user. The animal image includes the animal contour diagram and / or the reference checkerboard.

[0110] In one embodiment, the weight prediction module 303 is specifically configured to:

[0111] Perform image semantic segmentation on the animal image using the weight prediction model to obtain a plurality of target animal regions;

[0112] Obtain target image parameters corresponding to the target animal regions from the image parameters;

[0113] Calculate animal body shape features according to the target image parameters, construct a three-dimensional animal portrait according to the animal body shape features, and estimate the weight of the target animal according to the three-dimensional animal portrait.

[0114] The present invention provides an animal weight prediction device. By performing standardized shooting on a target animal to obtain an animal image, the gyroscope is used to control the shooting angle to ensure that the shooting angle meets the requirements and reduce image distortion caused by angle changes; an animal contour diagram and a reference checkerboard are used to assist animal positioning to ensure that the position of the captured animal image is accurate and the contour is clear. By precisely controlling the shooting process, key image parameters can be fully collected; by determining whether the animal image meets preset shooting conditions according to the image parameters of the animal image, the animal images can be screened according to the preset shooting conditions to ensure that only the images that meet the shooting conditions are used for subsequent weight prediction, thereby improving the accuracy of the prediction; by inputting the animal images and image parameters that meet the shooting conditions into a pre-trained weight prediction model, the weight prediction model can quickly and accurately predict the weight of the target animal according to the animal image and image parameters, effectively improving the accuracy of animal image and image parameter processing, reducing losses caused by weighing errors, and further improving the efficiency and accuracy of insurance claim settlement and loss assessment.

[0115] For the specific limitations of the animal weight prediction device, reference may be made to the limitations of the animal weight prediction method in the foregoing text, which will not be elaborated herein. Each module in the above animal weight prediction device can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory in the computer device in the form of software, so as to facilitate the processor to call and execute the operations corresponding to each of the above modules.

[0116] To solve the above technical problems, an embodiment of the present application further provides a computer device. For details, please refer to Figure 5 , Figure 5 which is the basic structural block diagram of the computer device in this embodiment.

[0117] The computer device 6 includes a memory 61, a processor 62, and a network interface 63 that are communicatively connected to each other through a system bus. It should be noted that only the computer device 6 with the memory 61, the processor 62, and the network interface 63 is shown in the figure. However, it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented. Among them, those skilled in the art of the present technology can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0118] The computer device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device can perform human-computer interaction with the user through a keyboard, a mouse, a remote control, a touchpad, or a voice control device.

[0119] The memory 61 includes at least one type of readable storage medium, which includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 61 may be an internal storage unit of the computer device 6, such as the hard disk or memory of the computer device 6. In other embodiments, the memory 61 may also be an external storage device of the computer device 6, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, FlashCard, etc. equipped on the computer device 6. Of course, the memory 61 may also include both the internal storage unit and the external storage device of the computer device 6. In this embodiment, the memory 61 is generally used to store the operating system and various application software installed on the computer device 6, such as computer-readable instructions of the animal weight prediction method. In addition, the memory 61 may also be used to temporarily store various types of data that have been output or will be output.

[0120] In some embodiments, the processor 62 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor 62 is generally used to control the overall operation of the computer device 6. In this embodiment, the processor 62 is used to run the computer-readable instructions stored in the memory 61 or process data, such as running the computer-readable instructions of the animal weight prediction method.

[0121] The network interface 63 may include a wireless network interface or a wired network interface, and this network interface 63 is generally used to establish a communication connection between the computer device 6 and other electronic devices.

[0122] This application also provides another implementation manner, that is, to provide a computer-readable storage medium storing computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to execute the steps of the animal weight prediction method as described above.

[0123] The computer device, computer-readable storage medium, and computer-readable instructions provided by the embodiments of the present application can, by executing, through a processor, to obtain pre-acquired animal images, and determine whether the animal images meet preset shooting conditions according to the image parameters of the animal images, screen the animal images according to the preset shooting conditions, ensure that only the images that meet the shooting conditions are used for subsequent weight prediction, thereby improving the accuracy of prediction; by inputting the animal images and image parameters that meet the shooting conditions into a pre-trained weight prediction model, the weight prediction model can quickly and accurately predict the weight of the target animal according to the animal images and image parameters, effectively improving the accuracy of processing the animal images and image parameters, reducing the losses caused by weighing errors, and further improving the efficiency and accuracy of insurance claim settlement and loss assessment.

[0124] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-described embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disc), and includes several instructions for causing a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present application.

[0125] Obviously, the above-described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The preferred embodiments of the present application are given in the drawings, but do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosed content of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or perform equivalent replacements for some of the technical features. Any equivalent structure directly or indirectly using the content of the specification and drawings of the present application in other related technical fields shall be equally within the scope of the patent protection of the present application.

[0126] The non-company software tools or components that appear in the embodiments of the present application are only for illustrative introduction and do not represent actual use.

Claims

1. A method for predicting animal weight, characterized in that: include: obtaining pre-acquired images of the animal; Acquiring image parameters of the animal image, and determining whether the animal image meets preset shooting conditions according to the image parameters; If the animal image meets the shooting conditions, the animal image and the image parameters are input into a pre-trained weight prediction model, and the weight prediction model is used to predict the target animal weight based on the animal image and the image parameters.

2. The method according to claim 1, characterized in that The method of using the weight prediction model to predict the weight of the target animal according to the animal image and the image parameters includes: Using the weight prediction model to perform image semantic segmentation on the animal image to obtain a plurality of target animal regions; Acquire target image parameters corresponding to the target animal area from the image parameters; The body shape features of the animal are calculated according to the target image parameters, a three-dimensional animal portrait is constructed according to the animal body shape features, and the weight of the target animal is estimated according to the three-dimensional animal portrait.

3. The method according to claim 1, characterized in that The shooting conditions include shooting angle conditions; the acquiring of image parameters of the animal image, and determining whether the animal image meets the preset shooting conditions according to the image parameters, include: Acquire a target shooting angle of the animal image as an image parameter, wherein the target shooting angle is acquired by a gyroscope; A preset shooting angle condition is obtained, and the target shooting angle is compared with the shooting angle condition. When the target shooting angle satisfies the shooting angle condition, it is determined that the animal image satisfies the shooting condition; otherwise, the shooting condition is not satisfied.

4. The method according to claim 1, characterized in that The shooting condition includes a contour coincidence condition; the acquiring image parameters of the animal image and determining whether the animal image meets the preset shooting condition according to the image parameters include: Get the animal outline image set on the shooting page; Extracting the target animal outline from the animal image, matching the animal outline image with the target animal outline, and obtaining an outline overlap area and an outline non-overlap area; Calculate the contour overlap ratio as an image parameter according to the contour overlap area and the contour non-overlap area; A preset contour coincidence condition is obtained, and when the contour coincidence rate satisfies the contour coincidence condition, it is determined that the animal image satisfies the shooting condition; otherwise, the shooting condition is not satisfied.

5. The method according to claim 1, characterized in that The shooting condition includes a chessboard overlap rate condition, and the animal image includes actual chessboard grids and a target animal; the acquiring of image parameters of the animal image, and determining whether the animal image meets the preset shooting condition according to the image parameters, includes: Get the reference chessboard set on the shooting page; Determining an actual chessboard from the animal image, wherein the actual chessboard is placed directly below the target animal; Matching the reference chessboard with the actual chessboard to obtain a target chessboard overlap rate as an image parameter; Obtain a preset chessboard overlap rate condition, compare the target chessboard overlap rate with the chessboard overlap rate condition, and when the target chessboard overlap rate meets the chessboard overlap rate condition, determine that the animal image meets the shooting condition; otherwise, the shooting condition is not met.

6. The method according to any one of claims 1 to 5, characterized in that Before obtaining the pre-collected animal image, the method further includes: In response to a start signal triggered by a user, the camera and the gyroscope are controlled to start and a shooting page is displayed, wherein the shooting page is provided with an animal outline, a reference chessboard, a shooting angle display area, a focus adjustment slider and a wide-angle switch; In response to a shooting signal triggered by a user, an animal image captured by the camera, a target shooting angle, a focal length multiple and a wide-angle state collected by the gyroscope are obtained, wherein the animal image includes the animal outline and / or the reference checkerboard.

7. The method according to claim 1, characterized in that After determining whether the animal image meets the preset shooting conditions according to the image parameters, the method further includes: If the animal image does not meet the shooting condition, acquiring abnormal parameters that do not meet the shooting condition; Prompt information is generated according to the abnormal parameters to prompt the user to re-capture the animal image until the re-captured animal image meets the shooting condition.

8. An animal weight prediction device, characterized in that: include: An image acquisition module, used to acquire pre-collected animal images; A parameter acquisition module, used to acquire image parameters of the animal image, and determine whether the animal image meets preset shooting conditions according to the image parameters; A weight prediction module is used to input the animal image and the image parameters into a pre-trained weight prediction model if the animal image meets the shooting conditions, and use the weight prediction model to predict the target animal weight based on the animal image and the image parameters.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the animal weight prediction method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the animal weight prediction method according to any one of claims 1 to 7 are implemented.

Images