A contactless live feature data analysis method, device, equipment and medium

Abstract

The present application relates to the technical field of image detection, and can be applied to the business system platform of financial technology, medical health and the like, and discloses a non-contact living body feature data analysis method, device, equipment and medium, comprising: non-contact collection of a three-dimensional space scene and a reference age in a space within a radius range around a target living body; extraction of a living body contour area of the target living body from the three-dimensional space scene, determination of a mass measurement reference value of the target living body according to the area size of the living body contour area; generation of a measurement point position guide mark of the target living body according to the mass measurement reference value; calculation of living body length data and living body height data of the target living body by using the measurement point position guide mark, identification of living body breed information of the target living body by using a living body breed database; and calculation of a living body feature data analysis result of the target living body according to the reference age, the living body length data, the living body height data and the living body breed information. The present application can improve the accuracy of non-contact living body feature data analysis.

Description

Technical Field

[0001] This invention relates to the field of image detection technology, and in particular to a non-contact method, apparatus, equipment and medium for analyzing live body feature data. Background Technology

[0002] In agricultural insurance coverage for live animals, measuring body length and estimating weight are crucial for assessing the value of the insured animal. Currently, the industry primarily employs two technical solutions: one involves manually measuring the body length and height with a measuring tape, then using empirical formulas to estimate weight; the other, explored by some insurance companies, is a "static photography + software calculation" model, where a side view of the animal is taken with a mobile phone, and image recognition software is used to perform the relevant calculations. Simultaneously, the industry has a real need for the authenticity and traceability of measurement data. Therefore, to meet the demands for efficient length and weight measurement and data traceability in agricultural insurance coverage for live animals, it is necessary to innovate and upgrade the length and weight measurement technologies for live animal coverage to improve the accuracy of contactless analysis of live animal characteristic data.

[0003] In live animal health care settings, veterinarians often need to obtain live animal weight data to develop precise medication plans or assess growth and health status. Currently, this is mostly done through manual collaboration: one or two people hold the restless animal still, while another person uses a measuring tape to measure the body length from nose to tail root and the height from hoof to acromion, then estimates the weight using empirical formulas. In some scenarios, static side-view photos of the live animal are taken with a mobile phone, and data is extracted and calculated using simple image processing software. However, live animal examinations are prone to stress and struggle, leading to measurement errors. Furthermore, the examination environment is often a farmhouse with complex lighting, resulting in a lot of background interference in the photos and poor software recognition stability, ultimately leading to low accuracy in non-contact live animal feature data analysis.

[0004] In online agricultural insurance underwriting or live animal pledge valuation, accurate calculation of the underlying asset value requires obtaining live animal weight data. Currently, two main methods are used: one relies on manual collaboration, where farmers, following platform guidance, work with 2-3 people to hold the animal still, measure its length and height with a measuring tape, and then estimate its weight using empirical formulas; the other, some platforms use static photography, where farmers take side-view photos of the live animal and upload them to the backend, where image recognition software calculates the relevant data. However, the agitation of live animals leads to errors in manual measurement data, and the variable lighting and complex backgrounds in pastures result in low success rates for static photo recognition, ultimately leading to low accuracy in non-contact live animal feature data analysis.

[0005] In existing technologies, there is no target mobile device to collect 3D scene data of live animals at the insurance site. Instead, people rely on carrying measuring tapes to the site and verbal feedback from farmers to determine the age. There is no target area segmentation step, and people determine the measurement benchmarks such as the nose and tail base by visual inspection. Traditional image recognition only takes static photos and cannot eliminate background interference that may cause benchmark positioning errors. There are no visual guidance markers, and people need to repeatedly adjust the position to align with the measurement points. Body length and height are recorded by manual reading, and live animal breed information depends on manual judgment. Weight estimation uses fixed empirical formulas and does not dynamically adjust parameters according to breed and age. This results in low measurement efficiency, benchmark positioning errors, and a single calculation parameter, leading to low accuracy in non-contact live animal feature data analysis. Summary of the Invention

[0006] This invention provides a non-contact method, apparatus, equipment, and medium for analyzing live biometric data, in order to solve the problem of low accuracy in non-contact live biometric data analysis.

[0007] Firstly, a non-contact method for analyzing live biometric data is provided, including: The three-dimensional spatial scene within a preset radius around the target living body is collected non-contactly, and the reference age of the target living body is obtained. Extract the live body contour region of the target live body from the three-dimensional spatial scene, and determine the mass measurement benchmark value of the target live body based on the area size of the live body contour region; Based on the aforementioned quality measurement benchmark values, guide markers for the measurement points of the target living organism are generated; The live body length and height data of the target live organism are calculated using the measurement point guidance markers, and the live organism species information of the target live organism is identified using a preset live organism species database. The reference age, the live body length data, the live body height data, and the live species information are analyzed to generate the live characteristic data analysis results of the target live animal.

[0008] Secondly, a contactless liveness feature data analysis device is provided, comprising: The three-dimensional spatial scene acquisition module is used to acquire the three-dimensional spatial scene within a preset radius around the target living body in a non-contact manner, and to obtain the reference age of the target living body; The mass measurement benchmark value acquisition module is used to extract the live body contour region of the target live body from the three-dimensional spatial scene, and determine the mass measurement benchmark value of the target live body based on the area size of the live body contour region; The measurement point guidance mark generation module is used to generate measurement point guidance marks for the target living body based on the mass measurement benchmark value. The live animal species information identification module is used to calculate the live length and height data of the target live animal using the measurement point guidance markers, and at the same time to identify the live animal species information of the target live animal using a preset live animal species database; The live animal feature data analysis result generation module is used to analyze the reference age, the live animal length data, the live animal height data, and the live animal species information to generate the live animal feature data analysis result of the target live animal.

[0009] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described contactless live feature data analysis method.

[0010] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the above-described contactless liveness feature data analysis method.

[0011] In the aforementioned solution implemented by the non-contact live animal feature data analysis method, device, equipment, and medium, a three-dimensional spatial scene within a preset radius around the target live animal can be collected non-contactly via a client, and the reference age of the target live animal can be obtained; the live animal contour region of the target live animal is extracted from the three-dimensional spatial scene, and the mass measurement benchmark value of the target live animal is determined based on the area of ​​the live animal contour region; measurement point guidance marks of the target live animal are generated based on the mass measurement benchmark value; the live animal length and height data of the target live animal are calculated using the measurement point guidance marks, and the live animal species information of the target live animal is identified using a preset live animal species database; analysis... The reference age, the live animal length data, the live animal height data, and the live animal breed information are used to generate the live animal characteristic data analysis results for the target live animal. In this invention, collecting the three-dimensional scene around the target live animal and estimating its age can eliminate background interference. Extracting the live animal contour area to establish a mass measurement benchmark can avoid measurement failure caused by the movement of the live animal. Generating measurement point guidance marks lowers the operation threshold and improves the efficiency of body size measurement. Combining the breed database to identify breed information, and then calculating the mass estimation results based on multiple parameters, it breaks through the limitations of traditional fixed experience coefficients, improves the weight estimation accuracy, and also realizes non-contact measurement to ensure operation safety, lays the foundation for data storage, and can solve the problem of low accuracy in non-contact live animal characteristic data analysis. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a schematic diagram of an application environment for a non-contact liveness feature data analysis method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a non-contact liveness feature data analysis method according to an embodiment of the present invention; Figure 3 yes Figure 2 A flowchart illustrating a specific implementation method of step S1; Figure 4 yes Figure 2 A flowchart illustrating a specific implementation method of step S2; Figure 5 This is a schematic diagram of a non-contact liveness feature data analysis device according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention; Figure 7 This is another structural schematic diagram of a computer device according to one embodiment of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] The contactless live feature data analysis method provided in this invention can be applied to, for example... Figure 1In this application environment, the client communicates with the server via a network. The server can non-contactly acquire a three-dimensional spatial scene within a preset radius around the target live organism through the client, and obtain the reference age of the target live organism; extract the live organism's outline region from the three-dimensional spatial scene, and determine the target live organism's mass measurement benchmark value based on the area of ​​the live organism's outline region; generate measurement point guidance marks for the target live organism based on the mass measurement benchmark value; calculate the target live organism's live body length and live body height data using the measurement point guidance marks, and simultaneously identify the target live organism's live organism species information using a preset live organism species database; analyze the reference age and the live body length data. The live animal's height data and breed information are used to generate live animal feature data analysis results for the target live animal. In this invention, collecting the three-dimensional scene around the target live animal and estimating its age can eliminate background interference. Extracting the live animal's contour area to establish a mass measurement benchmark can avoid measurement failures caused by the movement of the live animal. Generating measurement point guidance markers lowers the operational threshold and improves the efficiency of body size measurement. Combining breed database to identify breed information, and then calculating the mass prediction result based on multiple parameters, this invention breaks through the limitations of traditional fixed experience coefficients, improves the accuracy of weight estimation, and also achieves contactless measurement to ensure operational safety, laying the foundation for data storage. This can solve the problem of low accuracy in contactless live animal feature data analysis. The client can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be implemented using a separate server or a server cluster composed of multiple servers. The invention will be described in detail below through specific embodiments.

[0016] Please see Figure 2 As shown, Figure 2 A flowchart illustrating a non-contact liveness feature data analysis method provided in an embodiment of the present invention includes the following steps: S1. Non-contactly collect the three-dimensional spatial scene within a preset radius around the target living body, and obtain the reference age of the target living body.

[0017] In this embodiment of the invention, the target live animal refers to various live animals that require insurable length measurement and weight estimation, such as cattle and sheep to be insured in a breeding scenario; the preset radius range refers to a specific spatial range set around the target live animal, suitable for operators to use AR (Augmented Reality) devices to carry out three-dimensional scene capture and data collection. This range must ensure that the device can clearly capture the outline of the target live animal and its surrounding environment, while also taking into account the safety and ease of operation of the operators; the three-dimensional spatial scene refers to a three-dimensional virtual scene containing the target live animal and its real-time surrounding environment, captured and constructed in real time using SLAM (Simultaneous Localization and Mapping) technology of AR devices; the reference age refers to the estimated age information about the current growth stage of the target live animal provided by the breeder of the target live animal based on their own breeding records or breeding experience, or it can be obtained by a pre-trained analysis model recognizing the influence of the target live animal or images.

[0018] In this embodiment of the invention, reference is made to Figure 3 As shown, the non-contact acquisition of a three-dimensional spatial scene within a preset radius around the target living body, and the obtaining of the reference age of the target living body, includes: S31. Use a preset mobile device to perform real-time scene scanning of the space within a preset radius around the target living body to obtain three-dimensional point cloud data. S32. Based on the three-dimensional point cloud data, perform three-dimensional modeling of the scene-related region of the target living body to obtain the three-dimensional spatial scene of the target living body; S33. Input the reference age of the target living organism through the device interaction interface of the mobile device.

[0019] In detail, the preset mobile device refers to a mobile terminal with AR (Augmented Reality) functionality, capable of loading SLAM (Simultaneous Localization and Mapping) scene modeling plugins and target liveness detection models (such as lightweight deep learning target detection models), and able to run corresponding business applications. It must support GPS positioning to record the collection location and have image capture and video recording functions to obtain information about the target liveness and its surrounding environment. The 3D point cloud data refers to the massive set of 3D coordinate points of all object surfaces in the space captured by the device's SLAM technology when the preset mobile device performs a real-time scene scan of the space within a preset radius around the target liveness. After the operator operates the preset mobile device to open the corresponding business application, the device automatically loads the SLAM scene modeling plugin and the target liveness detection model, while obtaining the current location information (such as GPS positioning) and prompting the operator to input the estimated age of the target liveness. Subsequently, the device performs a real-time scene scan of the space within a preset radius around the target liveness, continuously capturing the surface features and spatial location information of the target liveness and its surrounding environment in the space through SLAM technology, and converting this information into 3D point cloud data containing massive 3D coordinate points.

[0020] Specifically, the scene-related region refers to the relevant area surrounding the target living body, including the target living body itself and its real-time environment, which may affect the accuracy of 3D modeling and subsequent measurement work. This region needs to cover the spatial range that can completely present the three-dimensional shape of the target living body, and at the same time include environmental elements that may interfere with the segmentation of the target living body's contour or need to be excluded. It is the core spatial range for subsequent 3D modeling. After the operator uses a preset mobile device to acquire the 3D point cloud data of the space within a preset radius around the target living body, the device processes the 3D point cloud data based on the loaded SLAM scene modeling plugin. First, it identifies and delineates the scene-related region, and then, combined with the target living body recognition model, it accurately segments the point cloud information of the target living body from the 3D point cloud data of the scene-related region, removes the point cloud data of irrelevant background in the region, and then, based on the segmented target living body point cloud information and the necessary environmental point cloud information retained in the scene-related region, it constructs a 3D spatial scene that can completely and clearly present the three-dimensional shape of the target living body and its real-time environment.

[0021] Furthermore, the device interaction interface refers to the application interface running on a preset mobile device with AR functionality, used for the underwriting length measurement and weight estimation of live animals. This interface can present operation guidance information, receive information input by the operator, and subsequently display AR guidance markers, measurement data, and other content. It is the core carrier for information interaction between the operator and the device. After the operator opens the business application for underwriting length measurement and weight estimation of live animals on the preset mobile device, the device will complete the loading of the SLAM scene modeling plugin and the target live animal recognition model, and obtain the current GPS location. The device interaction interface will then display a prompt message. Based on the breeding records or breeding experience of the target live animal, the operator enters the estimated age of the target live animal in the designated input area of ​​the interaction interface. After the input is completed, the device automatically associates and binds the reference age with the subsequently collected three-dimensional spatial scene data and measurement data of the target live animal.

[0022] For example, in a medical scenario for patient limb rehabilitation assessment, medical staff use a medical mobile device with a pre-set rehabilitation assessment AR program to perform real-time scene scanning of a suitable radius space around the patient's limb to be assessed, obtaining three-dimensional point cloud data including the limb and surrounding medical instruments, hospital bed, and other elements; then, based on the three-dimensional point cloud data, three-dimensional modeling is carried out on the scene-related area covering the patient's limb and surrounding key medical environment to generate a three-dimensional spatial scene that can clearly present the three-dimensional shape of the limb; medical staff enter the patient's reference age through the device's interactive interface, providing basic data for subsequent limb size measurement and rehabilitation progress analysis.

[0023] For example, in a financial scenario involving livestock asset-backed financing, bank loan officers use a mobile tablet pre-installed with an AR valuation program for livestock assets to scan a 2-3 meter radius space around the livestock to be mortgaged in real time, obtaining 3D point cloud data including the livestock's physique and surrounding pens. Based on this data, a 3D model is created of the scene-related area covering the livestock and the environment required for the mortgage assessment, generating a 3D spatial scene that clearly presents the livestock's three-dimensional form. The loan officer then enters the estimated age of the livestock through the tablet's interface, providing basic data support for subsequent mortgage asset value calculation and risk assessment.

[0024] S2. Extract the live body contour region of the target live body from the three-dimensional spatial scene, and determine the mass measurement benchmark value of the target live body based on the area size of the live body contour region.

[0025] In this embodiment of the invention, the living contour region refers to the precisely segmented area that can completely present the outline of the target living animal in the three-dimensional spatial scene where the target living animal is located, after automatically excluding non-target elements in the scene, constructed through AR-related technologies.

[0026] In this embodiment of the invention, reference is made to Figure 4 As shown, extracting the live contour region of the target live body from the three-dimensional spatial scene includes: S41. Identify the target positioning area of ​​the target living being in the three-dimensional spatial scene; S42. Filter out non-target living elements in the three-dimensional space scene according to the target positioning area, and generate the initial contour area of ​​the target living object; S43. Perform dynamic contour tracking on the initial contour region to obtain the live contour region of the target living body.

[0027] In detail, the target localization area refers to a specific area in the three-dimensional spatial scene where the target live animal is located, which is set up to accurately obtain the key parts of the live animal required for measurement. This area can be locked by AR visualization guidance markers. This area must correspond precisely to the actual key parts of the live animal to ensure the accuracy of the subsequent measurement benchmark. Identifying this area requires loading a SLAM scene modeling plugin and a deep learning target detection algorithm on an AR-enabled device to construct the three-dimensional spatial scene where the target live animal is located in real time.

[0028] Specifically, non-target live elements refer to all live animals in the 3D space scene where the target live animal is located, except for the target live animal that needs to be measured and estimated in length and weight. The initial contour region refers to the region that is initially extracted from the scene after filtering out non-target live elements in the 3D space scene. This region can roughly present the outline of the target live animal and has eliminated the main interference from non-target live animals. The algorithm identifies all live elements in the scene, distinguishes between target live animals and non-target live elements by combining the determined target positioning region, automatically masks or removes the image regions corresponding to non-target live elements in the scene, and initially extracts the initial contour region that can roughly present the outline of the target live animal from the remaining region. Furthermore, the boundary of non-target live elements can be confirmed with the help of the frame matching algorithm to ensure that the filtering operation does not affect the relevant regions of the target live animal.

[0029] Furthermore, after acquiring the initial contour region of the target live animal, the device captures continuous frame images of the three-dimensional spatial scene in real time. It uses an inter-frame matching algorithm to compare the key features of the initial contour region in adjacent frames and adjusts the contour range in real time to adapt to the dynamic movement or posture changes of the target live animal. At the same time, it combines a deep learning target detection algorithm to help confirm the regional boundary of the target live animal, avoiding deviations in contour tracking caused by slight changes in the background or local movements of the target live animal. Finally, it continuously locks and obtains a live contour region that can accurately reflect the real-time complete shape of the target live animal and support subsequent measurements of key parts.

[0030] For example, in a patient limb rehabilitation assessment scenario, medical staff use an AR-enabled device to construct a three-dimensional spatial scene of the limb to be assessed. The device loads a SLAM plugin and a deep learning target detection algorithm to generate visual guide markers for the olecranon of the elbow joint and the radial styloid process of the wrist joint. By image recognition, it is determined that the overlap between the markers and the actual parts is more than 90% to identify the target positioning area. Then, based on this area, non-target living elements such as the hands assisted by medical staff and other patient limbs in the surrounding area are filtered out to obtain the initial contour area of ​​the limb to be assessed. Afterwards, the device uses an inter-frame matching algorithm to compare the edge nodes and morphological features of the initial contour in consecutive frames, and tracks the contour changes during limb movement in real time, finally obtaining a living contour area of ​​the limb that can accurately measure the joint movement angle.

[0031] For example, in a bank customer's palm vein liveness detection account opening scenario, staff use an AR-enabled device to construct a three-dimensional spatial scene of the customer's palm. The device loads a SLAM plugin and a deep learning object detection algorithm to generate visual guide markers for the center point of the palm and the base of the five fingers. After determining that the overlap between the markers and the actual parts meets the standard through image recognition, the target positioning area is identified. Then, based on this area, non-target live elements such as the customer's other hand and the hands of surrounding people are filtered out to obtain the initial contour area of ​​the palm. Afterwards, the device uses an inter-frame matching algorithm to compare the edge features of the initial contour in consecutive frames and tracks the contour changes when the palm moves slightly in real time, finally obtaining a palm liveness contour area from which vein features can be accurately extracted.

[0032] In this embodiment of the invention, the area size refers to specific area data extracted from the live outline region of the target live animal that reflects its body shape characteristics. It may be the two-dimensional projection area of ​​the live outline from a specific viewpoint, or it may be the area of ​​the region enclosed by key parts within the outline. This area needs to be able to correlate with the body size of the live animal. The quality measurement benchmark value refers to the initial reference value calculated by using the extracted live outline region area size as the core basic parameter, combined with the target live animal's species, age, and other characteristic parameters, through a preset algorithm. This value is used for further accurate calculation of the live animal's quality and provides a basis for the final quality measurement result.

[0033] In this embodiment of the invention, determining the mass measurement benchmark value of the target living organism based on the area of ​​the living organism contour region includes: Extract the contour dimension parameters and contour area correction coefficient of the living body contour region; The mass characteristic value of the target living body is calculated based on the area of ​​the living body contour region, the contour dimension parameter, and the contour area correction coefficient. The quality characteristic value is adjusted according to the growth and development coefficient corresponding to the reference age of the target living organism to obtain the benchmark value for quality measurement of the target living organism.

[0034] In detail, the contour dimension parameter is a linear parameter extracted from the live contour region of the target live animal that reflects its key body shape characteristics, such as the straight-line distance and vertical distance between different key parts within the contour. The contour area correction coefficient is a coefficient set in conjunction with the species, age, and other characteristics of the target live animal, used to correct the deviation in contour area calculation and make it more consistent with the actual body shape. Using the key marker points of the live contour that have been locked in the AR interface, the distance between the marker points is automatically calculated to extract the contour dimension parameter. At the same time, the algorithm matches the live animal species database and combines it with the input live animal age information to call the preset correction coefficient data for the corresponding species and age, thereby completing the separate extraction of the contour dimension parameter and the contour area correction coefficient.

[0035] Specifically, the quality characteristic value refers to the core value calculated based on the body shape-related data of the target live animal, which can directly relate to and reflect its actual quality. It is a key reference value for subsequently determining the specific quality of the live animal. This process is implemented through a device with AR functionality. The device calls up the extracted live animal contour area size, contour dimension parameters, and contour area correction coefficient. At the same time, it matches the target live animal's breed database and age information through an algorithm. These parameters are input into a quality calculation model trained based on massive amounts of data from similar live animals. Algorithms such as multiple linear regression are used to correct the deviation of the area size combined with the correction coefficient, and the contour dimension parameters are associated to construct the correspondence between body shape and quality, thereby calculating and outputting the quality characteristic value of the target live animal.

[0036] Furthermore, the growth and development coefficient is set in conjunction with the growth and development patterns of the target live animal at different age stages. It reflects the growth maturity of the live animal at the current estimated age and thus affects its quality measurement results. This coefficient is determined based on a large amount of growth data of live animals of different species and age ranges. The AR-enabled device first obtains the reference age of the target live animal, and then uses an algorithm to call a basic database that stores the growth and development coefficients corresponding to different species and ages. It matches the growth and development coefficient corresponding to the reference age, and then performs a correlation operation between the previously calculated quality characteristic value of the target live animal and the growth and development coefficient. The quality characteristic value is adjusted to adapt to the current growth stage to obtain the benchmark value for quality measurement of the target live animal.

[0037] For example, in a child growth and development assessment scenario, medical staff use an AR-enabled device to construct a three-dimensional spatial scene of the child's entire body, extract the contour dimension parameters of the child's live contour region, and simultaneously match the contour area correction coefficient from a growth and development database based on the child's gender. The size of the contour region area, the extracted contour dimension parameters, and the correction coefficient are input into a quality calculation model, and the algorithm calculates the child's quality characteristic value. Then, based on the child's reference age, the corresponding growth and development coefficient is retrieved from the database, and the quality characteristic value is adjusted using this coefficient to obtain the child's weight quality measurement benchmark value.

[0038] For example, in a bank customer's palm vein liveness verification account opening scenario, staff use an AR-enabled device to construct a three-dimensional spatial scene of the customer's palm, extract the contour dimension parameters of the palm's liveness contour area, and match the contour area correction coefficient from a biometric database based on the customer's adult / adolescent identity. The size of the palm contour area, the extracted contour dimension parameters, and the correction coefficient are input into a liveness quality calculation model, and the algorithm calculates the quality characteristic value reflecting the palm tissue density. Then, based on the customer's reference age, the corresponding growth and development coefficient is called to adjust the quality characteristic value to obtain the quality measurement benchmark value for palm liveness verification, which is used to determine whether it is a real live palm to avoid the risk of unauthorized account opening.

[0039] S3. Generate measurement point guide marks for the target living body based on the quality measurement benchmark values.

[0040] In this embodiment of the invention, the measurement point guidance mark refers to a visual mark generated in the AR interface based on the target live body's quality measurement benchmark value. Its function is to guide the operator to align the mark with the key parts of the target live body related to quality assessment. At the same time, the device can automatically determine the degree of overlap between the mark point and the actual part of the live body. When the degree of overlap reaches the set standard, it can automatically calculate the distance of the corresponding part of the live body, thereby providing basic data support for the accurate estimation of the target live body's quality.

[0041] In this embodiment of the invention, generating the measurement point guidance markers for the target living organism based on the mass measurement benchmark value includes: Identify the target measurement location of the target living organism based on the aforementioned quality measurement benchmark value; Extract the spatial coordinate data of the target measurement site within the living body contour region; Based on the spatial coordinate data, measurement point guidance marks for the target living body are generated.

[0042] In detail, the target measurement site refers to a specific part of a live animal that can provide basic dimensional data for estimating the mass of the target live animal and is directly related to the quality assessment. It is usually a part that can be used to calculate key body shape dimensions such as body length, body height, and chest circumference. By combining the mass measurement benchmark values, the segmented target live animal area is analyzed in the AR interface to accurately locate those parts that can be measured to obtain dimensional data such as body length, body height, and chest circumference that are directly related to the mass estimation, thereby completing the identification of the target measurement site of the target live animal.

[0043] Specifically, spatial coordinate data refers to the digital data used to accurately represent the specific location of the target measurement part in the 3D scene of the target living body constructed using SLAM technology on AR devices. This data can directly provide a positional reference for subsequent calculations of the target living body's length, height, and other body dimensions related to quality estimation. By loading a SLAM scene modeling plugin and a lightweight deep learning target detection algorithm onto an AR-enabled device, the system can capture the location of the target living body in real time and construct a 3D scene. At the same time, the algorithm accurately segments the living body's outline area to eliminate background interference such as fences and other objects. Visual guide markers corresponding to the target measurement parts are generated on the AR interface. The operator moves the device to align the guide markers with the actual target measurement parts of the living body. The device automatically determines the overlap between the markers and the actual parts through image recognition. When the overlap reaches a set standard, the system automatically extracts the digital position information of these target measurement parts within the constructed 3D scene living body outline area, completing the extraction of spatial coordinate data. During the extraction process, parameters related to quality estimation, such as the species and age of the target living body, are simultaneously associated to ensure that the coordinate data is compatible with the subsequent quality assessment process.

[0044] Furthermore, after extracting the spatial coordinate data of the target measurement part, visual measurement point guidance marks corresponding to the target measurement part are generated in the AR interface based on these spatial coordinate data. These marks are directly mapped to the actual location of the target living body part corresponding to the spatial coordinate data in the three-dimensional scene. The device will automatically determine the overlap between the guidance marks and the actual target living body part through image recognition. When the overlap reaches the set standard, the generation of the measurement point guidance marks is completed, ensuring that the guidance marks can accurately point to the actual living body measurement part corresponding to the spatial coordinate data, providing accurate guidance for subsequent automatic calculation of relevant body shape dimension data.

[0045] S4. Calculate the live body length and height data of the target live organism using the measurement point guidance markers, and simultaneously identify the live organism species information of the target live organism using a preset live organism species database.

[0046] In this embodiment of the invention, the live body length data refers to the straight-line distance between two feature points automatically calculated by the device in the AR interface after aligning the measurement point guide markers with the feature points at the front of the head and the base of the tail of the target live body, respectively; the live body height data refers to the vertical distance between two points automatically calculated by the device in the AR interface after aligning the measurement point guide markers with the highest feature point on the back or shoulder of the target live body and the virtual ground point on the plane where the live body is standing. Both are core data of the target live body size automatically calculated and obtained based on the AR measurement point guide markers aligned with the key feature points of the live body.

[0047] In this embodiment of the invention, the step of calculating the live body length and height data of the target living organism using the measurement point guidance markers includes: Identify the degree of overlap between the measurement point guide marker and the target part of the target living organism; When the overlap is greater than the preset overlap, record the coordinate data of the point corresponding to the measurement point guide mark; The body length and height of the target living organism are calculated based on the coordinate data of the points.

[0048] In detail, the target body part refers to the key feature of a live organism that needs to be aligned with the AR measurement point guide markers when measuring its body size. This includes the prominent front part of the head and the base of the tail for calculating body length, and the highest point of the back or shoulders for calculating body height. The overlap rate refers to the degree of visual recognition matching between the measurement point guide markers generated in the AR interface and the actual target body part; it is the core indicator for determining whether the guide markers are accurately aligned with the target body part. Using an AR-enabled device, a 3D scene of the measurement site is first constructed in real time. Simultaneously, a deep learning object detection algorithm is used to segment the target live organism area to eliminate background interference. Then, measurement point guide markers for the key feature parts of the live organism pop up in the AR interface. The device uses image recognition technology to capture the visual matching between the guide markers and the actual target body part in real time, automatically determining the overlap rate between the two.

[0049] Specifically, the preset overlap refers to the overlap threshold set in advance by the system before performing AR-guided measurement on the target live body, used to determine whether the measurement point guide markers in the AR interface are accurately aligned with the target part of the live body; the point coordinate data refers to the specific position coordinate information of the guide marker in three-dimensional space generated by the device based on the real-time constructed three-dimensional scene of the scene when the measurement point guide marker is accurately aligned with the target part of the live body, including parameters that reflect the spatial position. When the measurement point guide marker corresponding to the key target part of the live body pops up in the AR interface, the device uses image recognition technology to capture and determine the overlap between the guide marker and the actual target part of the live body in real time. When the determination result shows that the overlap is greater than the preset overlap threshold set in advance by the system, the device automatically records the spatial coordinate information of the measurement point guide marker in the three-dimensional scene at this time, that is, the point coordinate data.

[0050] Furthermore, two sets of coordinate data for the protruding part of the head and the base of the tail are extracted for body length calculation. The straight-line distance between these two coordinate points in three-dimensional space is automatically calculated, and this distance is the live body length data of the target living body. At the same time, two sets of coordinate data for the highest part of the back or shoulder and the virtual point on the ground are extracted for body height calculation. The vertical distance between these two coordinate points in three-dimensional space is automatically calculated, and this distance is the live body height data of the target living body. The entire process is automatically completed based on the acquired coordinate data, without the need for manual intervention in the calculation process.

[0051] In this embodiment of the invention, the preset live animal breed database refers to a data set that the system has built and stored in advance, containing the breed classifications of various common live animals to be insured and the typical characteristics of each breed. This database covers the mainstream breed categories of different live animals. The live animal breed information refers to the breed category information of the target live animal determined by matching the appearance characteristics of the target live animal with the feature data in the preset live animal breed database.

[0052] In this embodiment of the invention, the step of identifying the live species information of the target live organism using a preset live species database includes: Extract the target appearance features of the target live organism, and calculate the feature matching degree between the target appearance features and the standard appearance features corresponding to each live organism in the preset live organism variety database; Candidate live organisms with a feature matching degree greater than a preset matching degree threshold are selected from the live organism variety database; The feature matching scores of the candidate live organisms are sorted in descending order, and the candidate live organism with the highest feature matching score is taken as the live organism species information of the target live organism.

[0053] In detail, target appearance features refer to the appearance attributes of the target live animal that can be captured through images, specifically including external features such as coat color and body shape that reflect the differences in the live animal breed. Standard appearance features refer to the appearance attributes that are representative of each recorded live animal breed in a pre-set live animal breed database, and are typical external features commonly possessed by that live animal breed. Feature matching degree refers to the quantified degree of similarity in attribute between the extracted target appearance features of the target live animal and the standard appearance features of a certain live animal breed in the pre-set live animal breed database, and is the core reference indicator for determining the breed to which the target live animal belongs. When performing AR measurement on a target live animal, the device with AR capability simultaneously captures its image. The target appearance features of the target live animal are extracted from the captured image by the live animal classification model of the AI ​​platform. At the same time, the AI ​​platform calls the pre-set live animal breed database, retrieves the standard appearance features corresponding to each live animal breed in the database, and then calculates the similarity between the extracted target appearance features and the standard appearance features of each live animal breed in the database through an algorithm.

[0054] Specifically, the preset matching degree threshold is a quantitative critical value set in advance by the system to determine whether the similarity between the target appearance features of the target live organism and the standard appearance features of a certain variety in the live organism variety database meets the standard. It is the core criterion for screening eligible varieties. Candidate live organisms refer to live varieties in the live organism variety database whose corresponding standard appearance features match the target appearance features of the target live organism with a feature matching degree greater than the preset matching degree threshold. They are the candidates for further determining the specific variety of the target live organism. After the AI ​​platform calculates the feature matching degree between the target appearance features of the target live organism and the standard appearance features of each live variety in the preset live organism variety database, it calls the preset matching degree threshold set in advance by the system. It compares the calculated feature matching degree of each variety with the threshold, and filters out all live varieties in the live organism variety database whose feature matching degree values ​​are greater than the preset matching degree threshold. These filtered varieties are the candidate live organisms.

[0055] Furthermore, based on the candidate live organisms previously selected from the live organism database and the feature matching degree corresponding to each candidate live organism, the feature matching degree values ​​of these candidate live organisms are arranged in descending order. The candidate live organism with the highest feature matching degree value after arrangement is directly determined as the live organism information of the target live organism to be identified.

[0056] S5. Analyze the reference age, the live body length data, the live body height data, and the live species information to generate the live characteristic data analysis results of the target live animal.

[0057] In this embodiment of the invention, the live animal feature data analysis result is an estimated weight of the live animal calculated based on a three-dimensional dynamic weight estimation model of "breed-age-measured dimension", such as health status characteristics and weight status characteristics. Specifically, the model first matches the corresponding breed coefficient with the live animal breed information, matches the corresponding growth and development coefficient with the reference age, and then uses these coefficients together with the measured live animal body length data and live animal body height data as input parameters. The model outputs the estimated weight of the live animal in real time through a multiple linear regression algorithm. Moreover, after adding a certain number of real live animal weight data, the model can automatically optimize the model parameters such as breed coefficient and growth and development coefficient through a gradient descent algorithm, thereby ensuring that the weight estimation accuracy of different types of live animals continues to improve.

[0058] In detail, this process relies on a three-dimensional dynamic weight estimation model based on "breed-age-measured dimension," which is trained and constructed using massive amounts of live animal data from different breeds and age ranges. The model first matches the corresponding breed coefficient from the basic database based on the target animal's breed information, and matches the corresponding growth and development coefficient based on the target animal's estimated age. Then, the matched breed coefficient, growth and development coefficient, and the measured target animal's body length and height data are used as input parameters and fed into a multiple linear regression algorithm for real-time calculation. The result is then output as the analysis of the target animal's live animal characteristic data. Furthermore, after adding 1,000 sets of actual weight data for the target animal, the model automatically optimizes the breed coefficient, growth and development coefficient, and other model parameters using a gradient descent algorithm to ensure continuous improvement in the accuracy of subsequent live animal weight estimation for other target animals.

[0059] Specifically, during the AR measurement process, the entire video, including AR interface guide lines and real-time display of measurement data, is automatically recorded. Simultaneously, measurement time, geographical location, and operator information are collected to generate a unique evidence package. This package is then uploaded to the blockchain system in real-time via API. This achieves two key benefits: firstly, the measurement process is replayable and the data is tamper-proof, mitigating the risk of data tampering during live animal insurance coverage and solving the problem of data authenticity verification; secondly, both farmers and insurance companies can query the measurement video and original data through blockchain nodes, improving data transparency and reducing underwriting disputes arising from questionable data authenticity. Furthermore, the evidence package undergoes MD5 value verification before transmission to prevent transmission corruption, ensuring data integrity and providing a reliable basis for subsequent claims verification and other business processes related to live animal insurance. This further reduces the underwriting and operational risks for insurance companies and supports the standardized development of live animal insurance business.

[0060] Furthermore, the AR interface generates visual guide lines and measurement marks, allowing operators to align key parts of the live animal with the guide lines without professional training. The device automatically calculates the body length and height, significantly lowering the operational threshold and reducing measurement errors caused by improper manual operation. It also eliminates the need for multiple people to assist in positioning, saving labor costs. Moreover, it supports multi-part data linkage acquisition, automatically associating the live animal species and binding the age information entered by the operator when measuring body length and height. This avoids mismatches between species, age, and measurement data that are prone to occur during manual recording, ensuring the relevance and accuracy of basic data. It also reduces the workload of manually recording multi-dimensional data, further improving overall measurement and data acquisition efficiency, making it suitable for the actual operational needs of various live animal insurance sites.

[0061] As can be seen, in the above scheme, collecting the three-dimensional scene around the target live animal and estimating its age can eliminate background interference, extracting the outline area of ​​the live animal to establish a mass measurement benchmark can avoid measurement failure caused by the movement of the live animal, generating measurement point guidance marks reduces the operation threshold and improves the efficiency of body size measurement, combining the variety database to identify variety information, and then calculating the mass estimation result based on multiple parameters, it breaks through the limitations of traditional fixed experience coefficients, improves the weight estimation accuracy, and also realizes non-contact measurement to ensure operation safety, lays the foundation for data storage, and can solve the problem of low accuracy of non-contact live animal feature data analysis.

[0062] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0063] In one embodiment, a contactless liveness feature data analysis device 100 is provided, which corresponds one-to-one with the contactless liveness feature data analysis method described in the above embodiments. For example... Figure 5 As shown, the contactless live animal feature data analysis device 100 includes a three-dimensional spatial scene acquisition module 101, a quality measurement benchmark value acquisition module 102, a measurement point guidance mark generation module 103, a live animal species information identification module 104, and a live animal feature data analysis result generation module 105. Detailed descriptions of each functional module are as follows: The three-dimensional spatial scene acquisition module 101 is used to acquire the three-dimensional spatial scene within a preset radius range around the target living body in a non-contact manner, and to obtain the reference age of the target living body; The mass measurement benchmark value acquisition module 102 is used to extract the live body contour region of the target live body from the three-dimensional space scene, and determine the mass measurement benchmark value of the target live body according to the area size of the live body contour region. The measurement point guide mark generation module 103 is used to generate measurement point guide marks for the target living body based on the mass measurement benchmark value. The live animal species information identification module 104 is used to calculate the live body length data and live body height data of the target live animal using the measurement point guidance markers, and at the same time to identify the live animal species information of the target live animal using a preset live animal species database; The live animal feature data analysis result generation module 105 is used to analyze the reference age, the live animal length data, the live animal height data and the live animal species information to generate the live animal feature data analysis result of the target live animal.

[0064] In one embodiment, the three-dimensional spatial scene acquisition module 101, when performing non-contact acquisition of the three-dimensional spatial scene within a preset radius around the target living body and obtaining the reference age of the target living body, is used for: Using a pre-set mobile device, a real-time scene scan is performed on the space within a pre-set radius around the target living body to obtain three-dimensional point cloud data; Based on the three-dimensional point cloud data, a three-dimensional model is performed on the scene-related region of the target living body to obtain the three-dimensional spatial scene of the target living body. The reference age of the target living organism is input through the device's interface.

[0065] In one embodiment, the quality measurement reference value acquisition module 102, when extracting the live contour region of the target live body from the three-dimensional spatial scene, is used to: Identify the target location region of the target living object in the three-dimensional spatial scene; Based on the target positioning region, non-target living elements in the three-dimensional spatial scene are filtered out to generate the initial contour region of the target living object; Dynamic contour tracking is performed on the initial contour region to obtain the live contour region of the target living body.

[0066] In one embodiment, the mass measurement reference value acquisition module 102, when performing the process of determining the mass measurement reference value of the target living body based on the area size of the living body contour region, is further configured to: Extract the contour dimension parameters and contour area correction coefficient of the living body contour region; The mass characteristic value of the target living body is calculated based on the area of ​​the living body contour region, the contour dimension parameter, and the contour area correction coefficient. The quality characteristic value is adjusted according to the growth and development coefficient corresponding to the reference age of the target living organism to obtain the benchmark value for quality measurement of the target living organism.

[0067] In one embodiment, the measurement point guidance mark generation module 103, when generating measurement point guidance marks for the target living body based on the mass measurement reference value, is used to: Identify the target measurement location of the target living organism based on the aforementioned quality measurement benchmark value; Extract the spatial coordinate data of the target measurement site within the living body contour region; Based on the spatial coordinate data, measurement point guidance marks for the target living body are generated.

[0068] In one embodiment, the live animal species information identification module 104, when performing the calculation of the live animal length data and live animal height data using the measurement point guidance markers, is used to: Identify the degree of overlap between the measurement point guide marker and the target part of the target living organism; When the overlap is greater than the preset overlap, record the coordinate data of the point corresponding to the measurement point guide mark; The body length and height of the target living organism are calculated based on the coordinate data of the points.

[0069] In one embodiment, the live animal species information identification module 104, when performing the task of identifying the live animal species information of the target live animal using a preset live animal species database, is further configured to: Extract the target appearance features of the target live organism, and calculate the feature matching degree between the target appearance features and the standard appearance features corresponding to each live organism in the preset live organism variety database; Candidate live organisms with a feature matching degree greater than a preset matching degree threshold are selected from the live organism variety database; The feature matching scores of the candidate live organisms are sorted in descending order, and the candidate live organism with the highest feature matching score is taken as the live organism species information of the target live organism.

[0070] This invention provides a non-contact live animal feature data analysis device. By collecting the three-dimensional scene around the target live animal and estimating its age, background interference can be eliminated. Extracting the live animal contour area to establish a mass measurement benchmark can avoid measurement failures caused by the movement of the live animal. Generating measurement point guidance marks lowers the operation threshold and improves the efficiency of body size measurement. Combining with a variety database to identify variety information, and then calculating the mass estimation result based on multiple parameters, it breaks through the limitations of traditional fixed empirical coefficients, improves the accuracy of weight estimation, and also achieves non-contact measurement to ensure operational safety and lays the foundation for data storage. It can solve the problem of low accuracy in non-contact live animal feature data analysis.

[0071] Specific limitations regarding the contactless liveness data analysis device can be found in the limitations of the contactless liveness data analysis method described above, and will not be repeated here. Each module in the aforementioned contactless liveness data analysis device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0072] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a contactless liveness feature data analysis method on the server side.

[0073] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 7 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a contactless liveness data analysis method on the client side.

[0074] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: The three-dimensional spatial scene within a preset radius around the target living body is collected non-contactly, and the reference age of the target living body is obtained. Extract the live body contour region of the target live body from the three-dimensional spatial scene, and determine the mass measurement benchmark value of the target live body based on the area size of the live body contour region; Based on the aforementioned quality measurement benchmark values, guide markers for the measurement points of the target living organism are generated; The live body length and height data of the target live organism are calculated using the measurement point guidance markers, and the live organism species information of the target live organism is identified using a preset live organism species database. The reference age, the live body length data, the live body height data, and the live species information are analyzed to generate the live characteristic data analysis results of the target live animal.

[0075] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: The three-dimensional spatial scene within a preset radius around the target living body is collected non-contactly, and the reference age of the target living body is obtained. Extract the live body contour region of the target live body from the three-dimensional spatial scene, and determine the mass measurement benchmark value of the target live body based on the area size of the live body contour region; Based on the aforementioned quality measurement benchmark values, guide markers for the measurement points of the target living organism are generated; The live body length and height data of the target live organism are calculated using the measurement point guidance markers, and the live organism species information of the target live organism is identified using a preset live organism species database. The reference age, the live body length data, the live body height data, and the live species information are analyzed to generate the live characteristic data analysis results of the target live animal.

[0076] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0077] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0078] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0079] It should be noted that if any software tools or components not belonging to our company appear in the embodiments of this application, they are merely for illustrative purposes and do not represent actual use.

[0080] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A non-contact method for analyzing live biometric data, characterized in that, include: The three-dimensional spatial scene within a preset radius around the target living body is collected non-contactly, and the reference age of the target living body is obtained. Extract the live body contour region of the target live body from the three-dimensional spatial scene, and determine the mass measurement benchmark value of the target live body based on the area size of the live body contour region; Based on the aforementioned quality measurement benchmark values, guide markers for the measurement points of the target living organism are generated; The live body length and height data of the target live organism are calculated using the measurement point guidance markers, and the live organism species information of the target live organism is identified using a preset live organism species database. The reference age, the live body length data, the live body height data, and the live species information are analyzed to generate the live characteristic data analysis results of the target live animal.

2. The non-contact live biometric data analysis method as described in claim 1, characterized in that, The non-contact acquisition of a three-dimensional spatial scene within a preset radius around the target living body, and the acquisition of the reference age of the target living body, includes: Using a pre-set mobile device, a real-time scene scan is performed on the space within a pre-set radius around the target living body to obtain three-dimensional point cloud data; Based on the three-dimensional point cloud data, a three-dimensional model is performed on the scene-related region of the target living body to obtain the three-dimensional spatial scene of the target living body. The reference age of the target living organism is input through the device's interface.

3. The non-contact live biometric data analysis method as described in claim 1, characterized in that, Extracting the live contour region of the target live body from the three-dimensional spatial scene includes: Identify the target location region of the target living object in the three-dimensional spatial scene; Based on the target positioning region, non-target living elements in the three-dimensional spatial scene are filtered out to generate the initial contour region of the target living object; Dynamic contour tracking is performed on the initial contour region to obtain the live contour region of the target living body.

4. The non-contact live biometric data analysis method as described in claim 1, characterized in that, The step of determining the mass measurement benchmark value of the target living organism based on the area of ​​the living organism contour region includes: Extract the contour dimension parameters and contour area correction coefficient of the living body contour region; The mass characteristic value of the target living body is calculated based on the area of ​​the living body contour region, the contour dimension parameter, and the contour area correction coefficient. The quality characteristic value is adjusted according to the reference age of the target living organism to obtain the benchmark value for quality measurement of the target living organism.

5. The non-contact liveness feature data analysis method as described in claim 1, characterized in that, The step of generating measurement point guidance marks for the target living organism based on the quality measurement benchmark value includes: Identify the target measurement location of the target living organism based on the aforementioned quality measurement benchmark value; Extract the spatial coordinate data of the target measurement site within the living body contour region; Based on the spatial coordinate data, measurement point guidance marks for the target living body are generated.

6. The non-contact liveness feature data analysis method as described in claim 1, characterized in that, The calculation of the target living organism's living length and height using the measurement point guidance markers includes: Identify the degree of overlap between the measurement point guide marker and the target part of the target living organism; When the overlap is greater than the preset overlap, record the coordinate data of the point corresponding to the measurement point guide mark; The body length and height of the target living organism are calculated based on the coordinate data of the points.

7. The non-contact live biometric data analysis method as described in claim 1, characterized in that, The step of identifying the live species information of the target live organism using a preset live species database includes: Extract the target appearance features of the target live organism, and calculate the feature matching degree between the target appearance features and the standard appearance features corresponding to each live organism in the preset live organism variety database; Candidate live organisms with a feature matching degree greater than a preset matching degree threshold are selected from the live organism variety database; The feature matching scores of the candidate live organisms are sorted in descending order, and the candidate live organism with the highest feature matching score is taken as the live organism species information of the target live organism.

8. A non-contact liveness feature data analysis device, characterized in that, include: The three-dimensional spatial scene acquisition module is used to acquire the three-dimensional spatial scene within a preset radius around the target living body in a non-contact manner, and to obtain the reference age of the target living body; The mass measurement benchmark value acquisition module is used to extract the live body contour region of the target live body from the three-dimensional spatial scene, and determine the mass measurement benchmark value of the target live body based on the area size of the live body contour region; The measurement point guidance mark generation module is used to generate measurement point guidance marks for the target living body based on the mass measurement benchmark value. The live animal species information identification module is used to calculate the live length and height data of the target live animal using the measurement point guidance markers, and at the same time to identify the live animal species information of the target live animal using a preset live animal species database; The live animal feature data analysis result generation module is used to analyze the reference age, the live animal length data, the live animal height data, and the live animal species information to generate the live animal feature data analysis result of the target live animal.

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, it implements the steps of the non-contact liveness feature data analysis method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the non-contact liveness feature data analysis method as described in any one of claims 1 to 7.

Images