Infrared automatic measurement method and system for multi-posture large animal body size

Abstract

The invention discloses an infrared automatic measurement method and system for a multi-posture large animal body size, and relates to the technical field of infrared automatic measurement, and the method comprises the steps: carrying out the multi-point arrangement of infrared sensors in a measurement region, and obtaining an infrared sensor array; capturing a posture change framework sequence; obtaining K reliable attitude change frames; obtaining Q authentication key capture point sets passing authentication; obtaining Q shoulder height parameters, Q individual length parameters and Q chest circumference parameters; and obtaining a target shoulder height parameter, a target body length parameter and a target chest circumference parameter, and taking the target shoulder height parameter, the target body length parameter and the target chest circumference parameter as body size measurement results of the target measurement object. The technical problem that in the prior art, the accuracy is low when infrared automatic measurement is conducted on the multi-posture large animal body size is solved, and the technical effect of improving the reliability of the measurement result is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of infrared automatic measurement, and in particular to an infrared automatic measurement method and system for large animal body dimensions in multiple postures. Background Art

[0002] Traditional animal body measurement methods usually rely on manual operation or mechanical measuring equipment with fixed posture. These methods have high requirements on the animal posture. The animal needs to be fixed or restrained during the measurement process, which can easily cause the animal to be nervous or behave abnormally, affecting the accuracy of the measurement results. In addition, manual measurement is inefficient and difficult to adapt to the needs of fast and batch body size data collection in large-scale animal farms. In recent years, non-contact measurement technology has gradually been applied to the field of animal body measurement. Among them, infrared sensors have become a potential alternative technology due to their high precision and environmental adaptability. However, existing measurement methods based on infrared sensors still have problems such as strong dependence on a single posture, one-sided data collection, and insufficient dynamic measurement accuracy.

[0003] The prior art has the technical problem of low accuracy when performing infrared automatic measurement of the body dimensions of large animals in multiple postures. Summary of the invention

[0004] The present application provides a method and system for automatic infrared measurement of large animal body dimensions in multiple postures, which are used to solve the technical problem of low accuracy in the prior art when automatic infrared measurement of large animal body dimensions in multiple postures is performed.

[0005] In view of the above problems, the present application provides a method and system for automatic infrared measurement of large animal body dimensions in multiple postures.

[0006] In a first aspect of the present application, a method for automatic infrared measurement of large animal body size in multiple postures is provided, the method comprising:

[0007] Arrange infrared sensors at multiple points in the measurement area to obtain an infrared sensor array;

[0008] Guiding the target measurement object to the measurement area, and using the infrared sensor array to capture a frame sequence of posture changes of the target measurement object;

[0009] According to a preset feature indicator set, a basic feature indicator set of the target measurement object is collected, and a measurement standard posture frame is mined according to the basic feature indicator set to obtain a measurement standard posture frame, and reliability authentication is performed on the posture change frame sequence based on the measurement standard posture framework to obtain K reliable posture change frames, where K is an integer greater than or equal to 1;

[0010] Traversing and extracting the key capture points of the K reliable posture change frames, obtaining K key capture point sets, and performing capture point distribution association authentication according to the K key capture point sets to obtain Q authenticated key capture point sets that pass authentication, wherein Q is a positive integer less than or equal to K;

[0011] Extract body size parameters according to the Q authentication key capture point sets to obtain Q shoulder height parameters, Q body length parameters and Q chest circumference parameters;

[0012] The Q shoulder height parameters, Q body length parameters and Q chest circumference parameters are traversed to perform parameter steady-state screening, and a target shoulder height parameter, a target body length parameter and a target chest circumference parameter are obtained, and the target shoulder height parameter, the target body length parameter and the target chest circumference parameter are used as the body size measurement result of the target measurement object.

[0013] The second aspect of the present application provides an infrared automatic measurement system for large animal body size in multiple postures, the system comprising:

[0014] An infrared sensor array acquisition module is used to deploy infrared sensors at multiple points in the measurement area to obtain an infrared sensor array;

[0015] A posture change frame sequence capturing module is used to guide the target measurement object to the measurement area and capture the posture change frame sequence of the target measurement object using the infrared sensor array;

[0016] A reliable posture change framework acquisition module is used to collect a basic feature indicator set of the target measurement object according to a preset feature indicator set, perform measurement standard posture framework mining according to the basic feature indicator set, obtain a measurement standard posture framework, and perform reliability authentication on the posture change framework sequence based on the measurement standard posture framework to obtain K reliable posture change frameworks, where K is an integer greater than or equal to 1;

[0017] A certification key capture point set acquisition module is used to traverse and extract the key capture points of the K reliable posture change frames to obtain K key capture point sets, and perform capture point distribution association authentication according to the K key capture point sets to obtain Q certified key capture point sets that pass the authentication, wherein Q is a positive integer less than or equal to K;

[0018] A body size parameter extraction module is used to extract body size parameters according to the Q authentication key capture point sets to obtain Q shoulder height parameters, Q body length parameters and Q chest circumference parameters;

[0019] The body size measurement result obtaining module is used to traverse the Q shoulder height parameters, Q body length parameters and Q chest circumference parameters to perform parameter steady-state screening, obtain target shoulder height parameters, target body length parameters and target chest circumference parameters, and use the target shoulder height parameters, target body length parameters and target chest circumference parameters as the body size measurement results of the target measurement object.

[0020] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0021] The present application arranges infrared sensors at multiple points in a measurement area to obtain an infrared sensor array, then guides the target measurement object to the measurement area, uses the infrared sensor array to capture the posture change frame sequence of the target measurement object, and then collects the basic feature indicator set of the target measurement object according to a preset feature indicator set, mines the measurement standard posture frame according to the basic feature indicator set, obtains the measurement standard posture frame, and performs reliability certification on the posture change frame sequence based on the measurement standard posture framework to obtain K reliable posture change frames, where K is an integer greater than or equal to 1, and then traverses and extracts the K reliable posture change frames. Key capture points, obtain K key capture point sets, and perform capture point distribution association authentication based on the K key capture point sets, obtain Q authenticated key capture point sets that have passed authentication, where Q is a positive integer less than or equal to K, extract body size parameters based on the Q authenticated key capture point sets, obtain Q shoulder height parameters, Q body length parameters, and Q chest circumference parameters, traverse the Q shoulder height parameters, Q body length parameters, and Q chest circumference parameters for parameter steady-state screening, obtain target shoulder height parameters, target body length parameters, and target chest circumference parameters, and use the target shoulder height parameters, target body length parameters, and target chest circumference parameters as the body size measurement results of the target measurement object. The technical effect of improving the accuracy and stability of infrared automatic measurement of large livestock body size in multi-posture dynamic scenes is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 A schematic flow chart of an automatic infrared measurement method for large animal body dimensions in multiple postures provided in an embodiment of the present application;

[0024] Figure 2 A schematic diagram of the structure of an infrared automatic measurement system for large animal dimensions in multiple postures provided in an embodiment of the present application.

[0025] Explanation of reference numerals: infrared sensor array acquisition module 11, posture change frame sequence capture module 12, reliable posture change frame acquisition module 13, authentication key capture point set acquisition module 14, body size parameter extraction module 15, body size measurement result acquisition module 16. DETAILED DESCRIPTION

[0026] The present application provides an infrared automatic measurement method and system for large animal body dimensions in multiple postures, which are used to solve the technical problem of low accuracy in the prior art when performing infrared automatic measurement of large animal body dimensions in multiple postures.

[0027] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0028] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0029] Embodiment 1, as Figure 1 As shown, the present application provides a multi-posture infrared automatic measurement method for large animal body size, wherein the method comprises:

[0030] S100: deploying infrared sensors at multiple points in the measurement area to obtain an infrared sensor array;

[0031] S200: guiding the target measurement object to the measurement area, and using the infrared sensor array to capture a frame sequence of posture changes of the target measurement object;

[0032] Furthermore, the target measurement object is guided to the measurement area, and the infrared sensor array is used to capture the posture change frame sequence of the target measurement object. In the embodiment of the present application, step S200 further includes:

[0033] Using a flexible fence to guide the target measurement object to the measurement area, activating the infrared sensor array to capture infrared depth information and contour data of the target measurement object in real time, extracting the data captured by each frame of the infrared sensor array, and obtaining an initial posture change frame sequence;

[0034] The initial posture change frame sequence is subjected to fuzzy abnormal frames and out-of-range frames elimination to obtain the posture change frame sequence.

[0035] In a possible embodiment, the measurement area is an area where the body size of the target object is measured. A plurality of infrared sensors are arranged in the measurement area, and these sensors cover different angles of the measurement area in a specific spatial arrangement to ensure that the posture information of the target object can be fully captured, thereby obtaining the infrared sensor array. The infrared sensor array is a system composed of a plurality of infrared sensors that can work together to capture three-dimensional data or infrared depth information of the target object.

[0036] Preferably, after determining the measurement area, the sensor positions are reasonably arranged according to the size and movement characteristics of the target object, so that the infrared sensor can fully cover the entire measurement area. The role of the infrared sensor array is to provide multi-angle data capture capabilities to avoid data loss due to the limitation of the viewing angle of a single sensor. This step lays the foundation for subsequent data collection and processing, ensuring that the captured target object posture information is complete and reliable.

[0037] In a possible embodiment, the target measurement object refers to an animal (such as a cow, a horse, etc.) whose body size needs to be measured. The posture change frame sequence is a posture data sequence sorted by time formed by continuously capturing infrared depth information and contour data of the target measurement object by the infrared sensor array, which is used to describe the dynamic posture change of the target object.

[0038] By guiding the target object to enter the measurement area, the infrared sensor array is activated to capture in real time and record the posture change information of the target object. By obtaining the posture change frame sequence, the technical effect of providing raw data support for subsequent analysis is achieved. By capturing the dynamic posture changes of the target object, it is ensured that it can adapt to the different postures of the animal, providing basic data for subsequent reliability certification and body size parameter extraction.

[0039] Preferably, the flexible fence is a device for guiding animals into a designated area, which can effectively avoid the stress response of the target object when entering the measurement area and ensure its natural movement. The blurred abnormal frame is a low-quality frame caused by motion blur or sensor error. The out-of-range frame removal is to remove frames in the captured data that exceed the measurement area or do not meet the standards.

[0040] Preferably, after the target measurement object is guided to the measurement area through the flexible fence, the infrared sensor array begins to capture its infrared depth information and contour data in real time to generate an initial posture change frame sequence. After the initial posture change frame sequence is generated, the fuzzy abnormal frames and out-of-range frames are eliminated to obtain a high-quality posture change frame sequence. This step improves the reliability and effectiveness of the data through real-time data acquisition and preprocessing, laying the foundation for subsequent posture frame authentication and parameter extraction.

[0041] S300: collecting a basic feature indicator set of the target measurement object according to a preset feature indicator set, mining a measurement standard posture framework according to the basic feature indicator set, obtaining a measurement standard posture framework, and performing reliability authentication on the posture change framework sequence based on the measurement standard posture framework to obtain K reliable posture change frameworks, where K is an integer greater than or equal to 1;

[0042] Furthermore, the preset characteristic indicator set includes type, variety and body shape.

[0043] Further, a basic feature indicator set of the target measurement object is collected according to a preset feature indicator set, a measurement standard posture framework is mined according to the basic feature indicator set to obtain a measurement standard posture framework, and reliability authentication is performed on the posture change frame sequence based on the measurement standard posture framework to obtain K reliable posture change frames. Step S300 of the embodiment of the present application also includes:

[0044] Collecting indicators of the target measurement object according to the preset characteristic indicator set to obtain the basic characteristic indicator set;

[0045] Using the basic feature index set as an index, performing measurement standard posture frame mining and matching to obtain a matching measurement posture frame set;

[0046] The matching measurement posture frame set is centrally screened to determine a target matching measurement posture frame, and the target matching measurement posture frame is used as the measurement standard posture frame.

[0047] Further, the matching measurement posture frame set is centrally screened to determine the target matching measurement posture frame, and the target matching measurement posture frame is used as the measurement standard posture frame. Step S300 of the embodiment of the present application further includes:

[0048] randomly extracting a first matching measurement pose frame from the matching measurement pose frame set;

[0049] Traversing and calculating the similarity between the matching measurement posture frame set and the first matching measurement posture frame to obtain a first matching similarity set;

[0050] randomly extracting a second matching measurement posture frame from the matching measurement posture frame set again;

[0051] Traversing and calculating the similarity between the matching measurement posture frame set and the second matching measurement posture frame to obtain a second matching similarity set;

[0052] Determine whether the mean of the first matching similarity set is greater than or equal to the mean of the second matching similarity set, and if so, use the second matching measurement posture frame as the stage matching measurement posture frame;

[0053] Randomly extracting the Nth matching measurement posture frame from the matching measurement posture frame set again;

[0054] Traversing and calculating the similarity between the matching measurement posture frame set and the Nth matching measurement posture frame to obtain an Nth matching similarity set;

[0055] When the mean of the Nth matching similarity set is greater than or equal to the mean of the stage matching measurement posture frame, and N is greater than or equal to the preset maximum number of screening times, the screening is stopped and the Nth matching measurement posture frame is used as the target matching measurement posture frame.

[0056] In one embodiment of the present application, the preset feature index set is an important parameter for distinguishing and describing the target measurement object, including type (such as cattle or horses), breed (such as dairy cows, beef cattle) and body type (such as young, adult). The basic feature index set reflects the basic situation of the target measurement object and is used to mine the measurement standard posture framework. The measurement standard posture framework is a posture framework matched and screened according to the basic feature indicators, which serves as a reference standard for the reliability certification of the posture change framework. The K reliable posture change frameworks are frameworks that meet the requirements in terms of outline.

[0057] Preferably, a basic feature indicator set (such as type, variety and body shape) of the target measurement object is collected according to the preset feature indicator set, and this is used as an index to perform matching mining in a large database to obtain a corresponding candidate measurement posture frame set, that is, the matching measurement posture frame set.

[0058] Then, the matching measurement posture frame set is centrally screened, and multiple rounds of random extraction and similarity calculation are used to gradually approach the optimal matching posture frame. When the mean of the screened similarity set reaches stability and the number of screening rounds exceeds a preset value, the final target matching measurement posture frame is determined as the target matching measurement posture frame, and the target matching measurement posture frame is used as the measurement standard posture frame to perform reliability authentication on the captured posture change frame sequence, and the K reliable posture change frames are screened out.

[0059] Optionally, a first matching measurement posture frame is randomly extracted from the matching measurement posture frame set. The cosine similarity calculation formula is used to traverse and calculate the similarity between the matching measurement posture frame set and the first matching measurement posture frame to obtain a first matching similarity set. The first matching similarity set reflects the similarity between the first matching measurement posture frame and the frames in the matching measurement posture frame set.

[0060] A second matching measurement posture frame is randomly extracted from the matching measurement posture frame set again, and the similarity between the matching measurement posture frame set and the second matching measurement posture frame is traversed and calculated to obtain a second matching similarity set. Then, it is determined whether the mean of the first matching similarity set is greater than or equal to the mean of the second matching similarity set. If so, it indicates that the second matching measurement posture frame is better than the first matching measurement posture frame. At this time, the second matching measurement posture frame is used as the stage matching measurement posture frame.

[0061] By analogy, the Nth matching measurement posture frame is randomly extracted from the matching measurement posture frame set again, and the similarity between the matching measurement posture frame set and the Nth matching measurement posture frame is traversed and calculated using the cosine similarity calculation formula to obtain the Nth matching similarity set. When the mean of the Nth matching similarity set is greater than or equal to the mean of the matching measurement posture frame at the stage, and N is greater than or equal to the preset maximum number of screening times (the maximum number of screening times pre-set by a person skilled in the art), the screening is stopped, and the Nth matching measurement posture frame is used as the target matching measurement posture frame.

[0062] That is to say, by gradually increasing the number of screening rounds N, similarity calculation and mean comparison are performed on the randomly extracted reference frames in each round. When the new mean is greater than or equal to the mean of the stage matching frame and the preset maximum number of screening times is reached, the screening is stopped and the Nth matching measurement posture frame obtained in the last screening is used as the target matching measurement posture frame. Through multiple rounds of random extraction and similarity optimization, the centralized screening process can efficiently eliminate candidate frames that do not meet the standards and ensure that the final target frame is highly matched with the target object features. This framework, as a measurement standard posture framework, directly affects the accuracy and efficiency of subsequent reliability certification.

[0063] S400: traverse and extract the key capture points of the K reliable posture change frames to obtain K key capture point sets, and perform capture point distribution association authentication according to the K key capture point sets to obtain Q authenticated key capture point sets that pass authentication, wherein Q is a positive integer less than or equal to K;

[0064] Further, the key capture points of the K reliable posture change frames are traversed and extracted to obtain K key capture point sets, and capture point distribution association authentication is performed according to the K key capture point sets to obtain Q authenticated key capture point sets that pass authentication. Step S400 of the embodiment of the present application further includes:

[0065] Pre-constructing a key capture point association relationship, wherein the key capture point association relationship includes that the angles of a line connecting the shoulder height point and the starting point of the body length and the line connecting the shoulder height point and the end point of the body length meet a preset angle threshold, and the error between a chest circumference point distribution fitting section and a preset fitting section is less than a preset fitting error threshold;

[0066] Respectively extract K shoulder height points, K body length starting points and K body length end points from the K key capture point sets;

[0067] Traverse and connect K shoulder height points and K individual length starting points to obtain K first straight lines;

[0068] Traverse and connect K shoulder height points and K body length end points to obtain K second straight lines;

[0069] Determine whether the angles between the K first straight lines and the K second straight lines meet the preset angle threshold in the key capture point association relationship. If so, the first authentication is passed, and M first authentication key capture point sets are obtained, where M is a positive integer less than or equal to K and greater than or equal to Q;

[0070] M chest circumference point sets from the M first authentication key capture point sets are extracted for second authentication to obtain Q authentication key capture point sets.

[0071] Furthermore, step S400 in the embodiment of the present application further includes:

[0072] Perform distribution fitting based on the M chest circumference point sets to obtain M chest circumference point distribution fitting sections;

[0073] Determine whether the error between the M chest circumference point distribution fitting section and the preset fitting section is less than the preset fitting error threshold in the key capture point association relationship. If so, the second authentication is passed and Q authentication key capture point sets are obtained.

[0074] In a possible embodiment, the key capture points are characteristic points describing the body size of the target measurement object, including shoulder height point, body length starting point, body length end point and chest circumference point. The shoulder height point, body length starting point, body length end point and chest circumference point set are extracted from K reliable posture change frames to form K key capture point sets respectively as basic data for subsequent authentication. The K key capture point sets are distributed and associated authenticated from two dimensions to improve the reliability of the capture points. Through a two-stage verification method of gradually screening the capture point set, it is ensured that the output key capture point set meets the requirements of the overall measurement.

[0075] Preferably, the key capture point association relationship is a rule determined by a technician in this field based on the geometric relationship and distribution characteristics between the key capture points to verify the spatial consistency of the key capture points. The key capture point association relationship includes that the angles of the line connecting the shoulder high point and the starting point of the body length and the line connecting the shoulder high point and the end point of the body length meet the preset angle threshold, and the error between the chest circumference point distribution fitting section and the preset fitting section is less than the preset fitting error threshold. The preset angle threshold refers to the angle range between the shoulder high point and the line connecting the starting point and the end point of the body length, which is used to measure the rationality of the posture. The preset fitting error threshold is the geometric error range between the chest circumference point distribution fitting section and the reference section, which is used to verify the accuracy of the chest circumference point distribution.

[0076] Traverse and connect K shoulder height points and K body length starting points to obtain K first straight lines. Traverse and connect K shoulder height points and K body length end points to obtain K second straight lines. Obtain the angles of the K first straight lines and the K second straight lines respectively, and then determine whether the angles meet the preset angle threshold in the key capture point association relationship. If so, the first authentication is passed, and M first authentication key capture point sets are obtained, where M is a positive integer less than or equal to K and greater than or equal to Q.

[0077] Furthermore, M chest circumference point sets are extracted from the M first authentication key capture point sets, and the ellipse section fitting is performed using the least squares method to obtain M chest circumference point distribution fitting sections. The cosine similarity of the fitted M chest circumference point distribution fitting sections and the preset fitting sections are respectively calculated to obtain M fitting errors. Then, it is determined whether the M fitting errors are less than the preset fitting error threshold in the key capture point association relationship. If so, the second authentication is passed, and Q authentication key capture point sets are obtained.

[0078] Through phased certification (first certification and second certification), the key capture point set is gradually screened and optimized to ensure that the output capture point set not only conforms to the geometric relationship but also meets specific distribution rules, thereby providing an accurate and reliable data basis for subsequent body size parameter extraction.

[0079] S500: extracting body size parameters according to the Q authentication key capture point sets to obtain Q shoulder height parameters, Q body length parameters and Q chest circumference parameters;

[0080] S600: Traversing the Q shoulder height parameters, Q body length parameters and Q chest circumference parameters to perform steady-state parameter screening, obtaining a target shoulder height parameter, a target body length parameter and a target chest circumference parameter, and using the target shoulder height parameter, the target body length parameter and the target chest circumference parameter as the body size measurement result of the target measurement object.

[0081] In a possible embodiment, Q shoulder height points of the Q authentication key capture point sets are extracted respectively, and the heights of the Q shoulder height points to the horizontal plane of the measurement area are used as the Q shoulder height parameters. Furthermore, Q individual length starting points and Q individual length end points of the Q authentication key capture point sets are extracted respectively, and the Q individual length starting points and Q individual length end points are connected respectively to obtain the connection length, thereby obtaining the Q individual length parameters.

[0082] Extract Q chest circumference point sets from the Q authentication key capture point sets respectively, perform ellipse section fitting using the least square method, and obtain Q chest circumference point distribution fitting sections. Obtain the Q major axes and Q minor axes of the Q chest circumference point distribution fitting sections, use the ellipse perimeter fitting formula, input the Q major axes and Q minor axes into the ellipse perimeter fitting formula, and obtain Q chest circumference parameters. The ellipse perimeter fitting formula is: ; is the chest circumference parameter, a is the major axis, and b is the minor axis.

[0083] The Q shoulder height parameters, Q body length parameters and Q chest circumference parameters are traversed to perform parameter mean calculation, and the target shoulder height parameter, target body length parameter and target chest circumference parameter are obtained according to the mean calculation result, and then the target shoulder height parameter, target body length parameter and target chest circumference parameter are used as the body measurement result of the target measurement object. By integrating the measurement data in multiple posture scenarios, high-precision measurement of body parameters can be achieved without the target measurement object maintaining a standard posture, thereby achieving a technical effect of improving the accuracy of the measurement result.

[0084] In summary, the embodiments of the present application have at least the following technical effects:

[0085] This application forms an infrared sensor array by arranging infrared sensors at multiple points, and captures the posture change frame sequence of the target measurement object from multiple angles, realizing dynamic measurement capabilities under free movement or non-fixed postures. The basic characteristic indicators of the target measurement object are collected in combination with a preset characteristic indicator set, and a measurement standard posture framework is mined based on matching and screening as a benchmark for reliability certification, thereby obtaining K reliable posture change frameworks. On this basis, by extracting key capture points and performing distribution association certification, Q sets of certification key capture points are further screened out to ensure the geometric consistency and distribution accuracy of the measurement data. Subsequently, shoulder height, body length and chest circumference parameters are extracted according to the certification results, and abnormal data are eliminated through steady-state screening, and finally the target shoulder height parameters, target body length parameters and target chest circumference parameters are obtained as measurement results. The technical effect of improving the accuracy of measurement and data stability is achieved.

[0086] Embodiment 2, based on the same inventive concept as the infrared automatic measurement method of large animal body size in multiple postures in the above embodiment, Figure 2 As shown, the present application provides an infrared automatic measurement system for large animal body size in multiple postures. The system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0087] The infrared sensor array acquisition module 11 is used to deploy infrared sensors at multiple points in the measurement area to obtain an infrared sensor array;

[0088] A posture change frame sequence capturing module 12 is used to guide the target measurement object to the measurement area and capture the posture change frame sequence of the target measurement object using the infrared sensor array;

[0089] A reliable posture change framework acquisition module 13 is used to collect a basic feature indicator set of the target measurement object according to a preset feature indicator set, perform measurement standard posture framework mining according to the basic feature indicator set, obtain a measurement standard posture framework, and perform reliability authentication on the posture change framework sequence based on the measurement standard posture framework to obtain K reliable posture change frameworks, where K is an integer greater than or equal to 1;

[0090] The authentication key capture point set acquisition module 14 is used to traverse and extract the key capture points of the K reliable posture change frames to obtain K key capture point sets, and perform capture point distribution association authentication according to the K key capture point sets to obtain Q authenticated key capture point sets, where Q is a positive integer less than or equal to K;

[0091] A body size parameter extraction module 15 is used to extract body size parameters according to the Q authentication key capture point sets to obtain Q shoulder height parameters, Q body length parameters and Q chest circumference parameters;

[0092] The body size measurement result obtaining module 16 is used to traverse the Q shoulder height parameters, Q body length parameters and Q chest circumference parameters to perform parameter steady-state screening, obtain target shoulder height parameters, target body length parameters and target chest circumference parameters, and use the target shoulder height parameters, target body length parameters and target chest circumference parameters as the body size measurement results of the target measurement object.

[0093] Furthermore, the posture change frame sequence capturing module 12 is used to perform the following steps:

[0094] Using a flexible fence to guide the target measurement object to the measurement area, activating the infrared sensor array to capture infrared depth information and contour data of the target measurement object in real time, extracting the data captured by each frame of the infrared sensor array, and obtaining an initial posture change frame sequence;

[0095] The initial posture change frame sequence is subjected to fuzzy abnormal frames and out-of-range frames elimination to obtain the posture change frame sequence.

[0096] Furthermore, the reliable posture change frame obtaining module 13 is used to perform the following steps:

[0097] Collecting indicators of the target measurement object according to the preset characteristic indicator set to obtain the basic characteristic indicator set;

[0098] Using the basic feature index set as an index, performing measurement standard posture frame mining and matching to obtain a matching measurement posture frame set;

[0099] The matching measurement posture frame set is centrally screened to determine a target matching measurement posture frame, and the target matching measurement posture frame is used as the measurement standard posture frame.

[0100] Furthermore, the reliable posture change frame obtaining module 13 is used to perform the following steps:

[0101] randomly extracting a first matching measurement pose frame from the matching measurement pose frame set;

[0102] Traversing and calculating the similarity between the matching measurement posture frame set and the first matching measurement posture frame to obtain a first matching similarity set;

[0103] randomly extracting a second matching measurement posture frame from the matching measurement posture frame set again;

[0104] Traversing and calculating the similarity between the matching measurement posture frame set and the second matching measurement posture frame to obtain a second matching similarity set;

[0105] Determine whether the mean of the first matching similarity set is greater than or equal to the mean of the second matching similarity set, and if so, use the second matching measurement posture frame as the stage matching measurement posture frame;

[0106] Randomly extracting the Nth matching measurement posture frame from the matching measurement posture frame set again;

[0107] Traversing and calculating the similarity between the matching measurement posture frame set and the Nth matching measurement posture frame to obtain an Nth matching similarity set;

[0108] When the mean of the Nth matching similarity set is greater than or equal to the mean of the stage matching measurement posture frame, and N is greater than or equal to the preset maximum number of screening times, the screening is stopped and the Nth matching measurement posture frame is used as the target matching measurement posture frame.

[0109] Furthermore, the preset characteristic indicator set includes type, variety and body shape.

[0110] Furthermore, the authentication key capture point set acquisition module 14 is used to perform the following steps:

[0111] Pre-constructing a key capture point association relationship, wherein the key capture point association relationship includes that the angles of a line connecting the shoulder height point and the starting point of the body length and the line connecting the shoulder height point and the end point of the body length meet a preset angle threshold, and the error between a chest circumference point distribution fitting section and a preset fitting section is less than a preset fitting error threshold;

[0112] Respectively extract K shoulder height points, K body length starting points and K body length end points from the K key capture point sets;

[0113] Traverse and connect K shoulder height points and K individual length starting points to obtain K first straight lines;

[0114] Traverse and connect K shoulder height points and K body length end points to obtain K second straight lines;

[0115] Determine whether the angles between the K first straight lines and the K second straight lines meet the preset angle threshold in the key capture point association relationship. If so, the first authentication is passed, and M first authentication key capture point sets are obtained, where M is a positive integer less than or equal to K and greater than or equal to Q;

[0116] M chest circumference point sets from the M first authentication key capture point sets are extracted for second authentication to obtain Q authentication key capture point sets.

[0117] Furthermore, the authentication key capture point set acquisition module 14 is used to perform the following steps:

[0118] Perform distribution fitting based on the M chest circumference point sets to obtain M chest circumference point distribution fitting sections;

[0119] Determine whether the error between the M chest circumference point distribution fitting section and the preset fitting section is less than the preset fitting error threshold in the key capture point association relationship. If so, the second authentication is passed and Q authentication key capture point sets are obtained.

[0120] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0121] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0122] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A multi-posture infrared automatic measurement method for large animal body size, characterized in that: The method comprises: Arrange infrared sensors at multiple points in the measurement area to obtain an infrared sensor array; Guiding the target measurement object to the measurement area, and using the infrared sensor array to capture a frame sequence of posture changes of the target measurement object; According to a preset feature indicator set, a basic feature indicator set of the target measurement object is collected, and a measurement standard posture frame is mined according to the basic feature indicator set to obtain a measurement standard posture frame, and reliability authentication is performed on the posture change frame sequence based on the measurement standard posture framework to obtain K reliable posture change frames, where K is an integer greater than or equal to 1; Traversing and extracting the key capture points of the K reliable posture change frames, obtaining K key capture point sets, and performing capture point distribution association authentication according to the K key capture point sets to obtain Q authenticated key capture point sets that pass authentication, wherein Q is a positive integer less than or equal to K; Extract body size parameters according to the Q authentication key capture point sets to obtain Q shoulder height parameters, Q body length parameters and Q chest circumference parameters; The Q shoulder height parameters, Q body length parameters and Q chest circumference parameters are traversed to perform parameter steady-state screening, and a target shoulder height parameter, a target body length parameter and a target chest circumference parameter are obtained, and the target shoulder height parameter, the target body length parameter and the target chest circumference parameter are used as the body size measurement result of the target measurement object.

2. The infrared automatic measurement method of large animal body size in multiple postures as claimed in claim 1, characterized in that: Guiding the target measurement object to the measurement area, and using the infrared sensor array to capture the posture change frame sequence of the target measurement object, including: Using a flexible fence to guide the target measurement object to the measurement area, activating the infrared sensor array to capture infrared depth information and contour data of the target measurement object in real time, extracting the data captured by each frame of the infrared sensor array, and obtaining an initial posture change frame sequence; The initial posture change frame sequence is subjected to fuzzy abnormal frames and out-of-range frames elimination to obtain the posture change frame sequence.

3. The infrared automatic measurement method of large animal body size in multiple postures as claimed in claim 1, characterized in that: According to a preset feature indicator set, a basic feature indicator set of the target measurement object is collected, and a measurement standard posture frame is mined according to the basic feature indicator set to obtain a measurement standard posture frame, and reliability authentication is performed on the posture change frame sequence based on the measurement standard posture framework to obtain K reliable posture change frames, including: Collecting indicators of the target measurement object according to the preset characteristic indicator set to obtain the basic characteristic indicator set; Using the basic feature index set as an index, performing measurement standard posture frame mining and matching to obtain a matching measurement posture frame set; The matching measurement posture frame set is centrally screened to determine a target matching measurement posture frame, and the target matching measurement posture frame is used as the measurement standard posture frame.

4. The infrared automatic measurement method for large animal body size in multiple postures as claimed in claim 3, characterized in that: Centrally screening the set of matching measurement posture frames to determine a target matching measurement posture frame, and using the target matching measurement posture frame as the measurement standard posture frame, including: randomly extracting a first matching measurement pose frame from the matching measurement pose frame set; Traversing and calculating the similarity between the matching measurement posture frame set and the first matching measurement posture frame to obtain a first matching similarity set; randomly extracting a second matching measurement posture frame from the matching measurement posture frame set again; Traversing and calculating the similarity between the matching measurement posture frame set and the second matching measurement posture frame to obtain a second matching similarity set; Determine whether the mean of the first matching similarity set is greater than or equal to the mean of the second matching similarity set, and if so, use the second matching measurement posture frame as the stage matching measurement posture frame; Randomly extracting the Nth matching measurement posture frame from the matching measurement posture frame set again; Traversing and calculating the similarity between the matching measurement posture frame set and the Nth matching measurement posture frame to obtain an Nth matching similarity set; When the mean of the Nth matching similarity set is greater than or equal to the mean of the stage matching measurement posture frame, and N is greater than or equal to the preset maximum number of screening times, the screening is stopped and the Nth matching measurement posture frame is used as the target matching measurement posture frame.

5. The infrared automatic measurement method for large animal body size in multiple postures as claimed in claim 4, characterized in that: The preset characteristic indicator set includes type, variety and body shape.

6. The infrared automatic measurement method for large animal body size in multiple postures as claimed in claim 1, characterized in that: Traversing and extracting the key capture points of the K reliable posture change frames, obtaining K key capture point sets, and performing capture point distribution association authentication according to the K key capture point sets to obtain Q authenticated key capture point sets that pass authentication, including: Pre-constructing a key capture point association relationship, wherein the key capture point association relationship includes that the angles of a line connecting the shoulder height point and the starting point of the body length and the line connecting the shoulder height point and the end point of the body length meet a preset angle threshold, and the error between a chest circumference point distribution fitting section and a preset fitting section is less than a preset fitting error threshold; Respectively extract K shoulder height points, K body length starting points and K body length end points from the K key capture point sets; Traverse and connect K shoulder height points and K individual length starting points to obtain K first straight lines; Traverse and connect K shoulder height points and K body length end points to obtain K second straight lines; Determine whether the angles between the K first straight lines and the K second straight lines meet the preset angle threshold in the key capture point association relationship. If so, the first authentication is passed, and M first authentication key capture point sets are obtained, where M is a positive integer less than or equal to K and greater than or equal to Q; M chest circumference point sets from the M first authentication key capture point sets are extracted for second authentication to obtain Q authentication key capture point sets.

7. The infrared automatic measurement method for large animal body size in multiple postures as claimed in claim 6, characterized in that: include: Perform distribution fitting based on the M chest circumference point sets to obtain M chest circumference point distribution fitting sections; Determine whether the error between the M chest circumference point distribution fitting section and the preset fitting section is less than the preset fitting error threshold in the key capture point association relationship. If so, the second authentication is passed and Q authentication key capture point sets are obtained.

8. An infrared automatic measurement system for large animal body size in multiple postures, characterized in that: The system is used to implement the infrared automatic measurement method of large animal body size in multiple postures according to any one of claims 1 to 7, and the system comprises: An infrared sensor array acquisition module is used to deploy infrared sensors at multiple points in the measurement area to obtain an infrared sensor array; A posture change frame sequence capturing module is used to guide the target measurement object to the measurement area and capture the posture change frame sequence of the target measurement object using the infrared sensor array; A reliable posture change framework acquisition module is used to collect a basic feature indicator set of the target measurement object according to a preset feature indicator set, perform measurement standard posture framework mining according to the basic feature indicator set, obtain a measurement standard posture framework, and perform reliability authentication on the posture change framework sequence based on the measurement standard posture framework to obtain K reliable posture change frameworks, where K is an integer greater than or equal to 1; A certification key capture point set acquisition module is used to traverse and extract the key capture points of the K reliable posture change frames to obtain K key capture point sets, and perform capture point distribution association authentication according to the K key capture point sets to obtain Q certified key capture point sets that pass the authentication, wherein Q is a positive integer less than or equal to K; A body size parameter extraction module is used to extract body size parameters according to the Q authentication key capture point sets to obtain Q shoulder height parameters, Q body length parameters and Q chest circumference parameters; The body size measurement result obtaining module is used to traverse the Q shoulder height parameters, Q body length parameters and Q chest circumference parameters to perform parameter steady-state screening, obtain target shoulder height parameters, target body length parameters and target chest circumference parameters, and use the target shoulder height parameters, target body length parameters and target chest circumference parameters as the body size measurement results of the target measurement object.

Images