Animal information processing method and apparatus, and network device

Abstract

The embodiment of the application provides an animal information processing method and device and network equipment, the method comprises: obtaining the first video data of the whole cycle of the animal, the first video data corresponds to the disease model of the animal; obtaining the first labeled data according to the first video data to train the animal information recognition model; the first labeled data includes basic information, behavior characteristics, disease model and biological index; obtaining test video data and inputting into the animal information recognition model for identification to obtain test information recognition result, the test information recognition result includes basic information, behavior characteristics, disease model and biological index; the behavior characteristics in the test information recognition result are analyzed to screen the action sequence with accuracy lower than the preset threshold, and after correction, the animal information recognition model is trained to determine the trained animal information recognition model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer information, in particular, the present application relates to an animal information processing method and device and network equipment. BACKGROUND

[0002] With the development of pose estimation and deep learning, in recent years, animal behavior analysis systems have been developed, mainly by extracting features from video for behavior analysis.

[0003] However, the existing scheme usually only analyzes the behavior of animals according to the video, and cannot obtain other information. SUMMARY

[0004] The embodiments of the present application provide an animal information processing method, device and network equipment, which can obtain video data of the whole cycle of animals, and obtain corresponding basic information, behavior characteristics, disease models, biological indicators, etc. as the annotation of video data to train the recognition model, so as to obtain the basic information, behavior characteristics, disease models and biological indicators of animals directly through the video by using the trained recognition model.

[0005] The technical solution is as follows:

[0006] In a first aspect, the present application provides an animal information processing method, which comprises: obtaining first video data of the whole cycle of animals, the first video data corresponding to a disease model of the animals, and the video data comprising a first video corresponding to spontaneous behavior of the animals and a second video corresponding to behavior test of the animals; obtaining first annotation data according to the first video data to train an animal information recognition model; the first annotation data comprising basic information, behavior characteristics, disease models and biological indicators; obtaining test video data and inputting it into the animal information recognition model for recognition to obtain test information recognition results, the test information recognition results comprising basic information, behavior characteristics, disease models and biological indicators; analyzing the behavior characteristics in the test information recognition results to screen action sequences with accuracy lower than a preset threshold, and training the animal information recognition model after correction to determine a trained animal information recognition model.

[0007] Further, before obtaining the video annotation data, the method further comprises: inputting the video data into a behavior analysis model to determine a behavior analysis result, and removing low-quality video from the video data according to the behavior analysis result; the method further comprises: obtaining video data to be analyzed and inputting it into the trained animal information recognition model to determine an analysis result of the animals, the analysis result comprising basic information, behavior characteristics, disease models and biological indicators.

[0008] Further, the step of training the animal information recognition model comprises: obtaining second video data of the animal and corresponding labels, wherein the second video data comprises videos of animals without diseases, and the second video data comprises third videos corresponding to spontaneous behaviors of the animals and fourth videos corresponding to behavior tests of the animals; obtaining the third videos and labels corresponding to the third videos, and training the animal information recognition model; obtaining the fourth videos and labels corresponding to the fourth videos, and training the animal information recognition model trained by the third videos; obtaining the first videos and labels corresponding to the first videos, and training the animal information recognition model trained by the fourth videos; and obtaining the second videos and labels corresponding to the second videos, and training the animal information recognition model trained by the first videos.

[0009] Further, the step of training the animal information recognition model comprises: obtaining fifth videos and labels corresponding to the fifth videos, wherein the labels corresponding to the fifth videos are obtained by analyzing the fifth videos; obtaining sixth videos and labels corresponding to the sixth videos, wherein the labels corresponding to the sixth videos are obtained by collecting data of the animal by using an auxiliary collecting device; training the animal information recognition model according to the fifth videos and the labels corresponding to the fifth videos, and training the animal information recognition model trained by the fifth videos according to the sixth videos and the labels corresponding to the sixth videos.

[0010] Further, the basic information comprises animal species, gender, age and weight; the behavior characteristics comprise behavior definition, duration and behavior parameters; the disease model comprises natural animals, disease model animals, disease course and main manifestations; and the biological indicators comprise heartbeat, blood pressure, respiratory rate and electroencephalogram.

[0011] Further, in the first video data of the whole cycle of the animal, the first video is recorded for 1 hour, and the second video is recorded for 5-15 minutes; the frequency of collecting the video of the animal with a life span of less than 3 years is once a week; the frequency of collecting the video of the animal with a life span of more than 3 years is once a month; the video recording is divided into single-view and multi-view; the multi-view comprises four peripheral cameras and one top camera; the single-view comprises the top camera; the basic information is collected when the first video data is obtained, and the first label data of the first video data is formed according to the basic information.

[0012] Further, the first label data further comprises sound information and environmental temperature information.

[0013] Further, the behavior characteristics in the test information recognition result are analyzed to screen the action sequence with an accuracy rate lower than a preset threshold, and after correction, the animal information recognition model is trained to determine the trained animal information recognition model, including: the behavior characteristics in the test information recognition result are analyzed in a manual screening manner to screen the action sequence with an accuracy rate lower than a preset threshold; the screened action sequence is filtered and corrected manually to determine test update data, and the animal information recognition model is adjusted according to the test update data to determine the trained animal information recognition model.

[0014] In a second aspect, the present application provides an animal information processing device, including: a video data acquisition module, configured to acquire first video data of an animal in a whole cycle, the first video data corresponding to a disease model of the animal, and the video data including first video corresponding to spontaneous behavior of the animal and second video corresponding to behavior test of the animal; an identification model training module, configured to acquire first labeling data according to the first video data to train an animal information recognition model; the first labeling data including basic information, behavior characteristics, a disease model, and biological indicators; a test data acquisition module, configured to acquire test video data and input the test video data into the animal information recognition model for recognition to obtain test information recognition result, the test information recognition result including basic information, behavior characteristics, a disease model, and biological indicators; and a test result analysis module, configured to analyze the behavior characteristics in the test information recognition result to screen an action sequence with an accuracy rate lower than a preset threshold, and after correction, train the animal information recognition model to determine a trained animal information recognition model.

[0015] In a third aspect, the present application provides a network device, including: a memory, a transceiver, and a processor; wherein the memory is configured to store a computer program; the transceiver is configured to transceive data under control of the processor; and the processor is configured to read the computer program in the memory and execute the method in the first aspect.

[0016] In a fourth aspect, the present application provides a storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the method in the first aspect.

[0017] The technical scheme provided by the present application has the beneficial effects that:

[0018] The scheme of the present application can be applied in an animal information identification scene, and basic information, behavior information and disease information of an animal can be obtained according to an animal video. Specifically, the scheme can obtain video data of the whole cycle of an animal, and obtain corresponding basic information, behavior characteristics, disease models and biological indexes as labels of the video data, so as to train an animal information identification model, and directly obtain the basic information, behavior characteristics, disease models and biological indexes of the animal by using the trained animal information identification model through the video, so as to realize the functions of obtaining basic information and disease diagnosis of the animal. Specifically, the scheme can obtain first video data of the whole cycle of an animal, the first video data corresponding to a disease model of the animal, and the video data including a first video corresponding to a spontaneous behavior of the animal and a second video corresponding to a behavior test of the animal; first label data is obtained according to the first video data, so as to train the animal information identification model; the first label data includes basic information, behavior characteristics, disease models and biological indexes; after the animal information identification model is trained, test video data can be obtained and input into the animal information identification model for identification, so as to obtain test information identification results, which include basic information, behavior characteristics, disease models and biological indexes; after the test information identification results are obtained, if all the test information identification results are corrected, the workload is large, therefore, the scheme can only analyze the behavior characteristics in the test information identification results, screen action sequences with an accuracy rate lower than a preset threshold, and train the animal information identification model according to the corrected test information after correction, so as to determine the trained animal information identification model. After the animal information identification model is trained, video data to be analyzed can be obtained and input into the trained animal information identification model, so as to determine an analysis result of the animal, which includes basic information, behavior characteristics, disease models and biological indexes. The scheme based on the video data can not only analyze the behavior of the animal, but also obtain other information of the animal, so that subsequent processing can be more convenient. In addition, the information collected by the scheme can be stored in a database, so as to be analyzed or used to train other models. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments of the present application will be briefly introduced.

[0020] Figure 1 is a flowchart of an animal information processing method according to an embodiment of the present application;

[0021] Figure 2 is a structural schematic diagram of an animal information processing device according to an embodiment of the present application;

[0022] Figure 3 is a structural block diagram of a network device according to an embodiment of the present application;

[0023] Figure 4 is a structural block diagram of a user equipment according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] Embodiments of the present application are described in detail below with reference to examples illustrated in the accompanying drawings, in which the same or similar elements or elements having the same or similar functions are denoted by the same or similar reference numerals throughout. The embodiments described below by reference to the accompanying drawings are exemplary and are for the purpose of explanation only, and are not to be construed as limiting the present application.

[0025] Those skilled in the art can understand that, unless specifically stated otherwise, the singular forms "a", "an", and "the" as used herein include plural forms, and "multiple" refers to two or more, and other quantifiers are similar. It should be further understood that the phrase "comprising" used in the specification of the present application means that the features, integers, steps, operations, elements, and / or components exist, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is "connected" or "coupled" to another element, it can be directly connected or coupled to the other element, or there can be intermediate elements. In addition, "connected" or "coupled" as used herein can include wireless connection or wireless coupling. The phrase "and / or" as used herein describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. The character " / " generally represents an "or" relationship between the associated objects before and after it.

[0026] The scheme of the present application can be applied in animal information identification scenarios, and basic information, behavior information, and disease information of animals can be obtained according to animal videos. Specifically, the scheme can obtain video data of the whole cycle of animals, and obtain corresponding basic information, behavior characteristics, disease models, biological indicators, etc. as annotations of the video data, to train an animal information identification model, so as to directly obtain the basic information, behavior characteristics, disease models, biological indicators of animals through the video by using the trained animal information identification model, to realize the functions of obtaining basic information and disease diagnosis of animals. In addition, the information collected by the scheme can also be stored in a database for analysis or training of other models.

[0027] The scheme of the present application is introduced as a whole below:

[0028] The present scheme aims to solve the problems of process redundancy and incomplete information in current animal behavior analysis. With the development of pose estimation and deep learning, animal behavior analysis systems have been developed in recent years, mainly by extracting features of body points of interest from videos for behavior analysis. For researchers in the biological field without computer foundation, this method has high learning cost, complex processing process, and the obtained information is limited to behavior information, and the accuracy needs to be improved. This makes the time and manpower cost of researchers not match the richness and accuracy of the obtained information. The present scheme can realize the direct acquisition of basic information, behavior information, disease information and biological indicators of animals through video based on the visual understanding technology of the full life cycle database, greatly reducing the time of researchers processing video and obtaining more target information.

[0029] The existing behavior analysis technology is mainly divided into two technical routes. One is to extract key points for behavior clustering analysis, and the clustering analysis model is Keypoint-MoSeq, Behavior Atlas, DeepLabCut, etc. Researchers obtain videos, mark key points manually, train models, supervise or unsupervise behavior clustering, and obtain behavior sequences after behavior annotation. A database based on this technology is established, which enables researchers to directly analyze videos to obtain behavior sequences. The second is to analyze behavior data through video understanding technology, and deep neural networks such as DeepEthogram use supervised machine learning networks to classify continuous behavior sequences from raw video streams.

[0030] The present scheme establishes a database based on visual understanding to reduce labor costs. First, the visual understanding method is used to reduce the operation difficulty by annotating through language. Second, the behavior videos of the life cycle of commonly used experimental animals are collected to train a multi-modal large model. This model can automatically identify the appearance characteristics of animals, reducing the tedious process of manual annotation. With the gradual enrichment of the database, the model can identify various commonly used experimental animals in most scenarios.

[0031] The present scheme enriches the full life cycle database to improve the universality of the method. The establishment of the full life cycle database of commonly used experimental animals can make the method applicable to most experimental animal application scenarios. In addition to spontaneous behavior, task behavior (such as behavior test) and carrying equipment scenarios can maintain high accuracy.

[0032] The present scheme uses high-quality videos analyzed by existing technologies to train the model to improve the accuracy. High-quality videos are selected into the database through preliminary analysis of videos by existing technologies Deeplabcut and Behaviour Atlas combined with manual assistance to ensure the accuracy of the model. The universality of the model is ensured by enriching the training content.

[0033] The scheme uses algorithms and devices to assist in obtaining comprehensive information. In addition to video information, the scheme integrates basic information of animals, behavioral characteristics, disease models, biological indicators, and other information into a database. With the assistance of related algorithms and auxiliary devices, researchers can obtain as much animal information as possible.

[0034] The basic content of the technical scheme of the scheme: the scheme uses a multi-modal large language model to analyze long-time behavior videos of common experimental animals throughout their life cycle, generating text descriptions of basic information, behavioral characteristics, disease models, and biological indicators of animals. The technical scheme includes data collection, model training, model adjustment, data analysis, and system construction.

[0035] Data collection: data collection animals include mice, rats, dogs, monkeys, and other commonly used animals and their disease models. Each disease model collects video of 10 or more animals throughout their life cycle. Spontaneous behavior recording time is 1 hour, and behavior test recording time is 5-15 minutes. Animals with a lifespan of less than 3 years are collected once a week, and animals with a lifespan of more than 3 years are collected once a month. Basic information such as body weight, heart rate, and respiratory rate is also collected. Video recording is divided into single-view and multi-view (principally multi-view), with multi-view consisting of 4 peripheral cameras and 1 top camera (single-view has only a top camera). Chessboard calibration is used during collection, and DeepLabCut and Behaviour Atlas are used for preliminary analysis after collection to exclude videos with chaotic behavior classification.

[0036] Model training: after excluding chaotic videos, reference the preliminary results of behavior analysis to label the video content with text. The labeled content includes four parts: basic information, behavioral characteristics, disease models, and biological indicators. Basic information includes animal species, gender, age, and body weight. Behavioral characteristics include behavior definition, duration, and behavior parameters. Disease models include natural animals or disease model animals, disease course, and main manifestations. Biological indicators include heart rate, blood pressure, respiratory rate, and EEG. For example: mouse, 8 weeks, 22g; rest for 30s, sniff for 10s, climb for 3s, look around for 4s, average speed is 20mm / s; anesthetized mouse, 15min after anesthesia, movement speed is less than 25mm / s; respiratory rate is 85 times / min, heart rate is 486 times / min, oxygen consumption is 1230mm2 / g live weight, ventilation is 14ml / min, tidal volume is 0.11ml, systolic pressure is 99mmHg, and diastolic pressure is 66mmHg.

[0037] Model adjustment: After training the model using the dataset, collect new videos as the test set and evaluate the model accuracy. By manually screening the action sequences with an accuracy of less than 80%, a short video dataset is formed, and filtering correction and manual correction are performed.

[0038] Data collection and model training are gradually added to the database in the order of natural animal spontaneous behavior, natural animal task behavior (including social behavior), disease model spontaneous behavior and animal behavior, and auxiliary devices (such as light recording, vital sign monitors, various invasive or non-invasive instruments, etc.).

[0039] System construction: Design a complete data analysis process, including basic information, behavior information, disease model, biological indicators, and their correlation and spatial feature distribution. Build a complete system to guide researchers to record videos, analyze videos, analyze data, and export results.

[0040] On the basis of the above embodiments, the embodiments of the present application also provide an animal information processing method, as shown in Figure 1 The method comprises the following steps:

[0041] Step 102, acquiring first video data of the whole cycle of the animal, the first video data corresponding to a disease model of the animal, the video data including a first video corresponding to spontaneous behavior of the animal and a second video corresponding to behavior test of the animal. As an optional embodiment, in the first video data of the whole cycle of the animal: the first video is recorded for 1 hour, and the second video is recorded for 5-15 minutes; the frequency of collecting videos of animals with a lifespan of less than 3 years is once a week; the frequency of collecting videos of animals with a lifespan of more than 3 years is once a month; video recording is divided into single-view and multi-view; multi-view is composed of 4 peripheral cameras and 1 top camera; single-view includes a top camera; basic information is collected when the first video data is acquired, and first annotation data of the first video data is formed according to the basic information. This scheme can use long-time videos to analyze basic information, behavior information, and disease and biological characteristics, and can reduce the content of annotation.

[0042] Step 104, acquiring first annotation data according to the first video data to train an animal information recognition model; the first annotation data includes basic information, behavior characteristics, disease model, and biological indicators. As an optional embodiment, the basic information includes animal species, gender, age, and weight; the behavior characteristics include behavior definition, duration, and behavior parameters; the disease model includes natural animals or disease model animals, disease course, and main performance; and the biological indicators include heart rate, blood pressure, respiratory rate, and electroencephalogram. As an optional embodiment, the first annotation data further includes sound information and environmental temperature information.

[0043] Step 106, acquire test video data and input into the animal information recognition model for recognition to obtain test information recognition results, the test information recognition results including basic information, behavior characteristics, disease model, biological indicators.

[0044] Step 108, analyze the behavior characteristics in the test information recognition results to screen the action sequence with an accuracy rate lower than a preset threshold, and after correction, train the animal information recognition model to determine the trained animal information recognition model.

[0045] The embodiments of the present application are similar to the embodiments of the above-mentioned embodiments, and the specific embodiments can refer to the specific embodiments of the above-mentioned embodiments, which will not be repeated here.

[0046] The scheme of the present application can be applied in the animal information recognition scene, and the basic information, behavior information and disease information of the animal can be acquired according to the animal video. Specifically, the present scheme can acquire the video data of the whole cycle of the animal, and acquire the corresponding basic information, behavior characteristics, disease model, biological indicators, etc. as the annotation of the video data, to train the animal information recognition model, so as to directly acquire the basic information, behavior characteristics, disease model, biological indicators of the animal through the video by using the trained animal information recognition model, to realize the functions of obtaining the basic information of the animal and disease diagnosis. Specifically, the present scheme can acquire the first video data of the whole cycle of the animal, the first video data corresponding to the disease model of the animal, the video data including the first video corresponding to the spontaneous behavior of the animal and the second video corresponding to the behavior test of the animal; according to the first video data, acquire the first annotation data to train the animal information recognition model; the first annotation data including: basic information, behavior characteristics, disease model, biological indicators; after training the animal information recognition model, test video data can be acquired and input into the animal information recognition model for recognition to obtain test information recognition results, the test information recognition results including basic information, behavior characteristics, disease model, biological indicators; after obtaining the test information recognition results, if all the test information recognition results are corrected, the workload is large, therefore the present scheme can only analyze the behavior characteristics in the test information recognition results, screen the action sequence with an accuracy rate lower than a preset threshold, and after correction, train the animal information recognition model according to the corrected test information to determine the trained animal information recognition model.

[0047] Before the video data is labeled, the scheme can input the video data into a behavior analysis model to analyze the behavior, so as to remove low-quality videos with animal behavior confusion. In addition, after the animal information recognition model is trained, the scheme can obtain video data to be analyzed and input the video data into the trained animal information recognition model to determine the analysis result of the animal, the analysis result including basic information, behavior characteristics, disease model and biological indicators. Based on the video data, the scheme can not only analyze the behavior of the animal, but also obtain other information of the animal, so that subsequent processing can be more convenient. Specifically, as an optional embodiment, before the video data is labeled, the method further includes: inputting the video data into a behavior analysis model to determine a behavior analysis result, and removing low-quality videos from the video data according to the behavior analysis result; the method further includes: obtaining video data to be analyzed and inputting the video data into the trained animal information recognition model to determine the analysis result of the animal, the analysis result including basic information, behavior characteristics, disease model and biological indicators. The training process of the model can include: inputting data into the model, the model obtaining an analysis result, and adjusting the model parameters according to the difference between the analysis result and the label, so as to obtain the trained model.

[0048] In the process of training the animal information recognition model, the scheme can preliminarily train the model according to the existing video and label of the disease-free animal, and then train the model according to the video and label of the animal with disease. Specifically, as an optional embodiment, the step of training the animal information recognition model includes: obtaining second video data of the animal and corresponding labels, the second video data including video of a disease-free animal, the second video data including a third video corresponding to a spontaneous behavior of the animal and a fourth video corresponding to a behavior test of the animal; obtaining the third video and the label corresponding to the third video, and training the animal information recognition model; obtaining the fourth video and the label corresponding to the fourth video, and training the animal information recognition model trained by the third video; obtaining the first video and the label corresponding to the first video, and training the animal information recognition model trained by the fourth video; obtaining the second video and the label corresponding to the second video, and training the animal information recognition model trained by the first video. According to the set order, the data is sequentially taken for model training, which can improve the recognition accuracy of the model.

[0049] The existing video data is labeled from different sources, some of which are manually labeled, some of which are obtained by video analysis, and some of which are obtained by acquisition equipment. Therefore, the present scheme can first train the model according to low-quality labels (such as inaccurate labels obtained by video analysis) and videos, and then fine-tune the model according to manual labels or labels obtained by acquisition equipment. Specifically, as an optional embodiment, the step of training the animal information recognition model comprises: obtaining a fifth video and a fifth video corresponding label, the fifth video corresponding label is obtained by analyzing the fifth video; obtaining a sixth video and a sixth video corresponding label, the sixth video corresponding label is obtained by the auxiliary acquisition device collecting data on the animal; training the animal information recognition model according to the fifth video and its label, and training the animal information recognition model trained by the fifth video according to the sixth video and its label.

[0050] After obtaining the test information recognition result, if all the test information recognition results are corrected, the workload is large, therefore the present scheme can only analyze the behavior characteristics in the test information recognition result, specifically, as an optional embodiment, the behavior characteristics in the test information recognition result are analyzed to screen the action sequence with an accuracy rate lower than a preset threshold, and after correction, the animal information recognition model is trained to determine the trained animal information recognition model, comprising: analyzing the behavior characteristics in the test information recognition result by manual screening to screen the action sequence with an accuracy rate lower than a preset threshold; filtering and correcting the screened action sequence to determine the test update data, and adjusting the animal information recognition model according to the test update data to determine the trained animal information recognition model.

[0051] On the basis of the above-mentioned embodiments, the present application further provides an animal information processing device, as shown in Figure 2 The device comprises:

[0052] The video data acquisition module 202 is configured to acquire first video data of the whole cycle of the animal, the first video data corresponding to a disease model of the animal, and the video data comprising a first video corresponding to a spontaneous behavior of the animal and a second video corresponding to a behavior test of the animal.

[0053] The recognition model training module 204 is configured to acquire first label data according to the first video data to train an animal information recognition model, and the first label data comprising basic information, behavior characteristics, a disease model, and biological indicators.

[0054] The test data acquisition module 206 is configured to acquire test video data and input the test video data into the animal information recognition model to obtain a test information recognition result, the test information recognition result including basic information, behavior characteristics, a disease model, and biological indexes.

[0055] The test result analysis module 208 is configured to analyze the behavior characteristics in the test information recognition result to screen a motion sequence with an accuracy rate lower than a preset threshold, and train the animal information recognition model after correction to determine a trained animal information recognition model.

[0056] The embodiments of the present application are similar to the embodiments of the above-described embodiments, and the specific embodiments can refer to the specific embodiments of the above-described embodiments, which will not be described herein.

[0057] The scheme of the present application can be applied in an animal information recognition scenario, and basic information, behavior information, and disease information of an animal can be acquired according to an animal video. Specifically, the scheme can acquire video data of an animal in a whole cycle, and acquire corresponding basic information, behavior characteristics, a disease model, and biological indexes as labels of the video data to train an animal information recognition model, so as to directly acquire the basic information, behavior characteristics, disease model, and biological indexes of the animal by using the trained animal information recognition model through the video to realize functions such as acquisition of basic information of the animal and disease diagnosis. Specifically, the scheme can acquire first video data of an animal in a whole cycle, the first video data corresponding to a disease model of the animal, the video data including a first video corresponding to a spontaneous behavior of the animal and a second video corresponding to a behavior test of the animal; first label data is acquired according to the first video data to train the animal information recognition model; the first label data includes basic information, behavior characteristics, a disease model, and biological indexes; after the animal information recognition model is trained, test video data can be acquired and input into the animal information recognition model to obtain a test information recognition result, the test information recognition result including basic information, behavior characteristics, a disease model, and biological indexes; after the test information recognition result is acquired, if all the test information recognition results are corrected, the workload is large, therefore, the scheme can only analyze behavior characteristics in the test information recognition result, screen a motion sequence with an accuracy rate lower than a preset threshold, and train the animal information recognition model according to the corrected test information after correction to determine a trained animal information recognition model. After the animal information recognition model is trained, video data to be analyzed can be acquired and input into the trained animal information recognition model to determine an analysis result of the animal, the analysis result including basic information, behavior characteristics, a disease model, and biological indexes. The scheme based on the video data can not only analyze the behavior of the animal, but also acquire other information of the animal, and subsequent processing can be more convenient.

[0058] It should be noted that the division of the units and / or modules in the embodiments of the present application is illustrative, and is only a logical function division. Actual implementation can have another division manner. In addition, each functional unit and / or module in each embodiment of the present application can be integrated in one processing unit and / or module, or each unit and / or module can be physically present alone, or two or more units and / or modules can be integrated in one unit and / or module. The integrated unit and / or module can be realized in the form of hardware or in the form of software functional unit and / or module.

[0059] When the integrated unit and / or module is realized in the form of software functional unit and / or module and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solutions of the present application or the part that essentially contributes to the related art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0060] In addition, the data transmission device and the data transmission method provided by the above embodiments are based on the same application concept. Since the principles of the method and the device for solving the problem are similar, the implementation of the device and the method can be referred to each other, and the repeated parts will not be described.

[0061] Figure 3 According to an exemplary embodiment, a structural block diagram of a network device is shown.

[0062] As shown in Figure 3 The network device 1100 at least includes a processor 1110, a memory 1120, and a transceiver 1130.

[0063] The transceiver 1130 is configured to receive and send data under the control of the processor 1110.

[0064] In Figure 3In this particular aspect, bus architecture can include any number of interconnected buses and bridges, specifically, various circuitry linking the one or more processors represented by processor 1110 and the memory represented by memory 1120. Bus architecture can also link various other circuitry such as peripheral devices, voltage regulators, and power management circuitry, which are well known in the art and thus, not further described herein. Bus interface provides an interface to the bus architecture. Transceiver 1130 can be a plurality of elements, i.e., including a transmitter and a receiver, providing a means and / or a module for communicating with various other apparatus over a transmission medium, including a wireless channel, a wired channel, optical cable, and the like.

[0065] Processor 1110 is responsible for managing the bus architecture and general processing, and memory 1120 can store data used by processor 1110 in executing operations.

[0066] Optionally, processor 1110 can be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or complex programmable logic device (CPLD), and processor 1110 can also take a multi-core architecture. Processor 1110 and memory 1120 can also be arranged physically separately.

[0067] Processor 1110, by invoking computer programs stored in memory 1120, is configured to execute any of the methods for allocating a cell radio network temporary identifier provided by the embodiments described above according to the executable instructions obtained.

[0068] Figure 4 A structural block diagram of a user equipment according to an exemplary embodiment is shown.

[0069] As shown in Figure 4 the user equipment 1300 at least includes processor 1310, memory 1320 and transceiver 1330.

[0070] Transceiver 1330 is configured to receive and send data under the control of processor 1310.

[0071] In Figure 4In one embodiment, the bus architecture can include any number of interconnecting buses and bridges, depending on the specific application of the processor 1310 and the memory 1320 that are linked together by the various circuits, which represent one or more processors for the processor 1310 and the memory for the memory 1320. The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and thus, will not be further described herein. The bus interface provides an interface. The transceiver 1330 can be a plurality of elements, i.e., including a transmitter and a receiver, providing a unit and / or module for communicating with various other apparatuses on transmission media, including wireless channels, wired channels, optical cables, etc. The user interface 1340 can also be an interface capable of connecting to the required devices for different user equipment, including but not limited to a keypad, a display, a speaker, a microphone, a joystick, etc.

[0072] The processor 1310 is responsible for managing the bus architecture and general processing, and the memory 1320 can store data used by the processor 1310 in performing operations.

[0073] Optionally, the processor 1310 can be a CPU (Central Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a CPLD (Complex Programmable Logic Device), and the processor 1310 can also adopt a multi-core architecture. The processor 1310 and the memory 1320 can also be physically arranged separately.

[0074] The processor 1310 is used to execute the executable instructions obtained according to the computer program stored in the memory 1320, to perform any one of the allocation methods of the cell radio network temporary identifier provided by the above-mentioned embodiments.

[0075] It should be noted that the above-mentioned apparatus provided by the embodiments of the present application can realize all the method steps realized by the above-mentioned method embodiments, and can achieve the same technical effects, and thus, the same parts and beneficial effects of the method embodiments will not be described in detail herein.

[0076] In addition, a storage medium is provided in the embodiments of the present application, and the storage medium stores a computer program. The computer program is executed by a processor to implement the data transmission method in the above embodiments. The storage medium can be any available medium or data storage device that the processor can access, including but not limited to a magnetic storage (for example, a floppy disk, a hard disk, a magnetic tape, a magneto-optical disk (MO), etc.), an optical storage (for example, a CD, a DVD, a BD, a HVD, etc.), and a semiconductor storage (for example, a ROM, an EPROM, an EEPROM, a non-volatile memory (NAND FLASH), a solid state disk (SSD), etc.), etc.

[0077] A program product is provided in the embodiments of the present application, for example, the program product is an FPGA chip or a DSP chip, and the program product includes executable instructions stored in a storage medium. The processor reads the executable instructions from the storage medium, so that the executable instructions are executed by the processor to implement the data transmission method in the above embodiments.

[0078] The scheme of the present application can be applied in an animal information identification scene, and basic information, behavior information and disease information of an animal can be obtained according to an animal video. Specifically, the scheme can obtain video data of the whole cycle of an animal, and obtain corresponding basic information, behavior characteristics, disease models, biological indicators and the like as labels of the video data, so as to train an animal information identification model, and directly obtain the basic information, behavior characteristics, disease models and biological indicators of the animal through the video by using the trained animal information identification model, so as to realize functions such as obtaining basic information and disease diagnosis of the animal. Specifically, the scheme can obtain first video data of the whole cycle of an animal, the first video data corresponding to a disease model of the animal, and the video data including a first video corresponding to a spontaneous behavior of the animal and a second video corresponding to a behavior test of the animal; first label data is obtained according to the first video data, so as to train the animal information identification model; the first label data includes basic information, behavior characteristics, disease models and biological indicators; after the animal information identification model is trained, test video data can be obtained and input into the animal information identification model for identification, so as to obtain test information identification results, the test information identification results including basic information, behavior characteristics, disease models and biological indicators; after the test information identification results are obtained, if all the test information identification results are corrected, the workload is large, therefore, the scheme can only analyze the behavior characteristics in the test information identification results, screen action sequences with an accuracy rate lower than a preset threshold, and train the animal information identification model according to the corrected test information after correction, so as to determine the trained animal information identification model. After the animal information identification model is trained, video data to be analyzed can be obtained and input into the trained animal information identification model, so as to determine an analysis result of the animal, the analysis result including basic information, behavior characteristics, disease models and biological indicators. The scheme based on the video data can not only analyze the behavior of the animal, but also obtain other information of the animal, and subsequent processing can be more convenient.

[0079] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems, or computer program products. Accordingly, the present application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application can be embodied in the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, magnetic disk storage and optical storage) having computer-usable program code embodied thereon.

[0080] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-executable instructions. These computer-executable instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0081] These processor-executable instructions may also be stored in a processor-readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the processor-readable memory produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0082] These processors can execute instructions that can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable device for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0083] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0084] The above description is only a partial embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. An animal information processing method characterized by comprising: The method comprises: acquiring first video data of the whole cycle of the animal, the first video data corresponding to a disease model of the animal, the video data comprising a first video corresponding to spontaneous behavior of the animal and a second video corresponding to behavior testing of the animal; acquiring first annotation data according to the first video data to train an animal information recognition model; the first annotation data comprising basic information, behavior characteristics, a disease model and biological indicators; acquiring test video data and inputting the test video data into the animal information recognition model to obtain a test information recognition result, the test information recognition result comprising basic information, behavior characteristics, a disease model and biological indicators; analyzing the behavior characteristics in the test information recognition result to screen a motion sequence with an accuracy lower than a preset threshold, and training the animal information recognition model after correction to determine a trained animal information recognition model; the step of training the animal information recognition model comprises: acquiring second video data of the animal and corresponding annotations, the second video data comprising a video of a disease-free animal, the second video data comprising a third video corresponding to spontaneous behavior of the animal and a fourth video corresponding to behavior testing of the animal; acquiring the third video and annotations corresponding to the third video and training the animal information recognition model; acquiring the fourth video and annotations corresponding to the fourth video and training the animal information recognition model trained by the third video; acquiring the first video and annotations corresponding to the first video and training the animal information recognition model trained by the fourth video; acquiring the second video and annotations corresponding to the second video and training the animal information recognition model trained by the first video.

2. The method of claim 1, wherein, Before acquiring the video annotation data, the method further comprises: inputting the video data into a behavior analysis model to determine a behavior analysis result and removing low-quality video from the video data according to the behavior analysis result; the method further comprises: acquiring video data to be analyzed and inputting the video data into the trained animal information recognition model to determine an analysis result of the animal, the analysis result comprising basic information, behavior characteristics, a disease model and biological indicators.

3. The method of claim 1, wherein, the step of training the animal information recognition model comprises: acquiring a fifth video and annotations corresponding to the fifth video, the annotations corresponding to the fifth video being obtained by analyzing the fifth video; acquiring a sixth video and annotations corresponding to the sixth video, the annotations corresponding to the sixth video being obtained by data acquisition of the animal by an auxiliary acquisition device; training the animal information recognition model according to the fifth video and the annotations thereof and training the animal information recognition model trained by the fifth video according to the sixth video and the annotations thereof.

4. The method of claim 1, wherein, the basic information comprises animal species, gender, age and weight; the behavior characteristics comprise behavior definition, duration and behavior parameters; the disease model comprises a natural animal or a disease model animal, a disease course and main manifestations; and the biological indicators comprise heartbeat, blood pressure, respiratory rate and electroencephalogram.

5. The method of claim 1, wherein, The first video data of the whole cycle of the animal: the first video recording time is 1 hour, and the second video recording time is 5-15 minutes; the frequency of collecting the video of the animal with a life span of less than 3 years is once a week; the frequency of collecting the video of the animal with a life span of more than 3 years is once a month; the video recording is divided into single view and multi-view; the multi-view is composed of 4 peripheral cameras and 1 top camera; The single view includes the top camera; the basic information is collected when the first video data is obtained, and the first annotation data of the first video data is formed according to the basic information.

6. The method of claim 1, wherein, The first annotation data further includes sound information and environmental temperature information.

7. The method of claim 1, wherein, The behavior characteristics in the test information recognition result are analyzed to screen the action sequence with an accuracy rate lower than a preset threshold, and after correction, the animal information recognition model is trained to determine the trained animal information recognition model, including: The behavior characteristics in the test information recognition result are analyzed to screen the action sequence with an accuracy rate lower than a preset threshold by means of artificial screening; The screened action sequence is filtered and corrected manually to determine the test update data, and the animal information recognition model is adjusted according to the test update data to determine the trained animal information recognition model. 8.An animal information processing apparatus, characterized by comprising: The device includes: A video data acquisition module is configured to acquire first video data of the whole cycle of an animal, wherein the first video data corresponds to a disease model of the animal, and the video data includes first video corresponding to spontaneous behavior of the animal and second video corresponding to behavior test of the animal; An identification model training module is configured to acquire first annotation data according to the first video data to train an animal information recognition model, wherein the first annotation data includes basic information, behavior characteristics, a disease model, and biological indicators; A test data acquisition module is configured to acquire test video data and input the test video data into the animal information recognition model to obtain test information recognition result, wherein the test information recognition result includes basic information, behavior characteristics, a disease model, and biological indicators; A test result analysis module is configured to analyze behavior characteristics in the test information recognition result to screen action sequences with an accuracy rate lower than a preset threshold, and train the animal information recognition model after correction to determine a trained animal information recognition model; The steps of training the animal information recognition model include: Acquiring second video data of the animal and corresponding annotations, wherein the second video data includes video of an animal without disease, and the second video data includes third video corresponding to spontaneous behavior of the animal and fourth video corresponding to behavior test of the animal; Acquiring the third video and annotations corresponding to the third video, and training the animal information recognition model; Acquiring the fourth video and annotations corresponding to the fourth video, and training the animal information recognition model trained by the third video; Acquiring the first video and annotations corresponding to the first video, and training the animal information recognition model trained by the fourth video; Acquiring the second video and annotations corresponding to the second video, and training the animal information recognition model trained by the first video.

9. A network device, comprising: ​ A memory, a transceiver, and a processor; wherein the memory is configured to store a computer program; the transceiver is configured to transceive data under the control of the processor; The processor is configured to read the computer program in the memory and execute the method as claimed in claims 1-7.

Images