CN115620378A

CN115620378A - Multi-view cow face intelligent acquisition method, device and system and related equipment

Info

Publication number: CN115620378A
Application number: CN202211411330.0A
Authority: CN
Inventors: 张锦华
Original assignee: Ping An Property and Casualty Insurance Company of China Ltd
Current assignee: Ping An Property and Casualty Insurance Company of China Ltd
Priority date: 2022-11-11
Filing date: 2022-11-11
Publication date: 2023-01-17

Abstract

The embodiment of the application belongs to the field of artificial intelligence, and relates to a multi-view cow face intelligent acquisition method, which comprises the following steps: acquiring video stream image data comprising a cow face and preprocessing the video stream image data; inputting the preprocessed video stream image data into a trained cattle face detection model to obtain a cattle face detection result, wherein the cattle face detection result comprises a plurality of frames of cattle face images at each visual angle, a cattle face detection frame of each frame of cattle face image and cattle face key information; and screening out multi-view cattle face images meeting preset quality standards from the cattle face images at all the view angles of the multi-frame according to the cattle face detection frame and the key information of the cattle face, and storing the multi-view cattle face images. This application detects the ox face image under multiframe each visual angle from video stream image data through the good ox face detection model of training in advance, then screens the ox face image according to predetermined quality standard, obtains the high-quality multi-view angle ox face image that accords with the standard to supply the ox face to discern and use, improve the accuracy and the robustness of ox face discernment.

Description

Multi-view cow face intelligent acquisition method, device and system and related equipment

Technical Field

The application relates to the technical field of artificial intelligence, in particular to a multi-view cow face intelligent acquisition method, device, system, computer equipment and readable storage medium.

Background

In the process of insurance application or claim settlement of cattle, accurate identification of the identity of cattle is one of the most important steps. The traditional method is to perform identification of cattle by ear tagging or implanting a biochip on each cattle, but the method is relatively high in cost and has a relatively high fraud risk. Therefore, the method for identifying the identity based on the biological information (the face characteristics) of the cattle is gradually popularized and used, has the advantages of convenience in use, high accuracy, low cost, low fraud risk and the like, and provides a quick, convenient, accurate and powerful means for the identity confirmation of the cattle in the processes of insurance verification and claim settlement and exploration. The cow face image snapshot collection is an upstream task for carrying out cow face identification, so that research and development of a high-precision and high-frame-rate cow face snapshot collection system is an important part of the system. And the ox face has strong stereoscopic impression, three faces are provided, key points of each face are inconsistent, and many current ox face recognition algorithms only use the ox face features of the front view angle for comparison, the ox face features of the side view angles (such as left and right view angles) are not collected and utilized, the situation that key points are blocked, the collected image quality is low and the like possibly exists in only one face is detected, the follow-up ox face recognition can also be influenced, and the accuracy and the robustness of the ox face recognition are not favorably improved.

Disclosure of Invention

The embodiment of the application aims to provide a multi-view cow face intelligent acquisition method, which can acquire high-quality multi-view cow face images for cow face identification and improve the accuracy and robustness of cow face identification.

In order to solve the above technical problem, an embodiment of the present application provides a multi-view bovine face intelligent acquisition method, which adopts the following technical scheme:

the multi-view cow face intelligent acquisition method comprises the following steps:

acquiring video stream image data comprising a cow face and preprocessing the video stream image data;

inputting the preprocessed video stream image data into a trained cattle face detection model to obtain a cattle face detection result, wherein the cattle face detection result comprises a plurality of frames of cattle face images at each visual angle, a cattle face detection frame of each frame of cattle face image and cattle face key information;

and screening out multi-view cattle face images meeting preset quality standards from the cattle face images at all the view angles of the multiple frames according to the cattle face detection frame and the key information of the cattle face, and storing the multi-view cattle face images.

Further, according to the ox face detection frame and the key information of the ox face, the multi-view ox face image which accords with the preset quality standard is screened from the ox face images at all the view angles of the plurality of frames, and the multi-view ox face image comprises:

and screening out a multi-view cattle face image set with the proportion larger than a preset proportion threshold value from the cattle face images at all the view angles of the multiple frames according to the proportion of the cattle face detection frame in the frame of the acquisition equipment.

Further, according to the key information of the cow face detection frame and the cow face, the step of screening out the multi-view cow face image meeting the preset quality standard from the cow face images at each view angle of the multi-frame comprises:

determining a cattle face plane relation based on the cattle face visual angle type and the cattle face key point information in the cattle face key information, and screening the multi-visual-angle cattle face image set according to the cattle face plane relation to obtain the cattle face images under multiple visual angles according with the quality standard.

Further, after the step of screening the multi-view bovine face image set according to the bovine face plane relationship, the method further includes:

and further screening the multi-view cattle face image set according to the number of the cattle face detection frames of each frame of cattle face image and the positions of the cattle face detection frames to obtain the cattle face images at multiple views according with the quality standard.

Further, the acquiring and preprocessing the video stream image data including the cow face includes:

performing frame rate conversion on the acquired cattle face video stream data through image resampling;

and performing noise reduction, enhancement and normalization processing on the cattle face video stream data after the frame rate conversion.

Further, the training step of the cattle face detection model comprises:

constructing a backbone network layer of the cattle face detection model by using ResNet50, and constructing a neck network layer of the cattle face detection model by using a convolution kernel and a deconvolution kernel;

the method for constructing the head network layer of the cattle face detection model according to the preset classification and/or regression tasks comprises the following steps: based on basic convolution, an activation function and bottleneck convolution, respectively constructing corresponding task branch networks to form the head network layer according to at least a preset thermodynamic diagram classification task of a cattle face central point, an offset regression task of the cattle face central point, a width and height regression task of a cattle face detection frame, a thermodynamic diagram classification task of the cattle face key points and an offset regression task of the cattle face key points;

the method comprises the steps of collecting cattle face data and conducting data processing to obtain a cattle face data set, using the cattle face data set to train and test a cattle face detection model to obtain a trained cattle face detection model.

In order to solve the above technical problem, an embodiment of the present application further provides a multi-view cow face intelligent acquisition device, which adopts the following technical scheme:

many visual angles ox face intelligent acquisition device includes:

the acquisition module is used for acquiring and preprocessing video stream image data comprising the cattle face;

the detection module is used for inputting the preprocessed video stream image data into a pre-trained cattle face detection model to obtain a cattle face detection result, wherein the cattle face detection result comprises a plurality of frames of cattle face images at each visual angle, a cattle face detection frame of each frame of cattle face image and cattle face key information;

and the screening module is used for screening out multi-view cattle face images meeting the preset quality standard from the cattle face images at all the view angles of the multi-frame according to the cattle face detection frame and the key information of the cattle face and storing the multi-view cattle face images.

In order to solve the technical problem, an embodiment of the present application further provides a multi-view bovine face intelligent acquisition system, including a video stream image input and preprocessing module, a multi-view bovine face intelligent detection module, an intelligent feedback module and a storage module, wherein the multi-view bovine face intelligent detection module is used for detecting and screening out high-quality multi-view bovine face images from the video stream images, the intelligent feedback module is used for carrying out feedback and prompt of acquisition operations to a video acquisition end and/or a user end according to the detected multi-view bovine face images, and the storage module is used for locally and/or remotely storing the acquired multi-view bovine face images.

In order to solve the above technical problem, an embodiment of the present application further provides a computer device, which adopts the following technical solutions:

the computer device comprises a memory and a processor, wherein computer readable instructions are stored in the memory, and the processor executes the computer readable instructions to realize the steps of the multi-view cow face intelligent acquisition method.

In order to solve the above technical problem, an embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solutions:

the computer readable storage medium stores computer readable instructions, and the computer readable instructions, when executed by the processor, implement the steps of the multi-view bovine face intelligent acquisition method.

Compared with the prior art, the embodiment of the application mainly has the following beneficial effects: the method comprises the steps of acquiring video streaming image data of a cow face, carrying out preprocessing such as data enhancement and normalization, inputting a pre-trained cow face detection model, detecting cow face images of multiple frames at all visual angles from the video streaming image data through the cow face detection model, detecting a cow face detection frame of each frame of cow face images and corresponding cow face key information, screening the cow face images of the multiple frames at all visual angles according to a preset quality standard based on the cow face detection frame and the corresponding cow face key information, obtaining high-quality multi-visual-angle cow face images meeting the standard, and using the cow face images for cow face identification, so that the accuracy and robustness of cow face identification are improved. The cattle face detection model in the embodiment of the invention is a single-stage detection network, and can simultaneously perform tasks of multi-face (left, middle and right faces) cattle face detection and 5 cattle face key point positioning, wherein the prediction of the cattle face key points is based on thermodynamic diagrams, so that the target cattle face detection model can more accurately position the key points, and the algorithm has good stability and high efficiency.

Drawings

In order to more clearly illustrate the solution of the present application, the drawings needed for describing the embodiments of the present application will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and that other drawings can be obtained by those skilled in the art without inventive effort.

FIG. 1 is an exemplary system architecture diagram in which the present application may be applied;

FIG. 2 is a flow diagram of one embodiment of a multi-view bovine face smart acquisition method according to the present application;

FIG. 3 is a schematic diagram of the overall structure of a cattle face detection model constructed according to the present application;

FIG. 4 is a schematic diagram of a 5-branch network structure of a head network layer of a cattle face detection model according to the present application;

FIG. 5 is a schematic structural diagram of one embodiment of a multi-view bovine face intelligent acquisition device according to the present application;

fig. 6 is a schematic structural diagram of an embodiment of a multi-view bovine face intelligent acquisition system according to the present application;

FIG. 7 is a schematic block diagram of one embodiment of a computer device according to the present application.

Detailed Description

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "including" and "having," and any variations thereof, in the description and claims of this application and the description of the above figures are intended to cover non-exclusive inclusions. The terms "first," "second," and the like in the description and claims of this application or in the above-described drawings are used for distinguishing between different objects and not for describing a particular order.

Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.

In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

As shown in fig. 1, fig. 1 is an exemplary system architecture diagram in which the present application may be applied, and system architecture 100 may include

terminal devices

101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the

terminal devices

101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.

The user may use the

terminal devices

101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The

terminal devices

101, 102, 103 may have various communication client applications installed thereon, such as a camera/video application, a web browser application, a shopping application, a search application, an instant messaging tool, a mailbox client, social platform software, and the like.

The

terminal devices

101, 102, 103 may be various camera devices having a display screen and supporting APP, web browsing, including but not limited to a camera, a video recorder, a smart phone with camera function, a tablet computer, a portable computer, a desktop computer, and the like.

The server 105 may be a server providing various services, such as a background server providing support for pages displayed on the

terminal devices

101, 102, 103.

It should be noted that the multi-view bovine face intelligent acquisition method provided by the embodiment of the present application is generally executed by a server/terminal device, and accordingly, the multi-view bovine face intelligent acquisition apparatus is generally disposed in the server/terminal device, and then transmits the acquired bovine face video stream image data to the server for processing through the terminal device and a network.

It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

With continuing reference to fig. 2, a flow diagram of one embodiment of a multi-view bovine face smart acquisition method according to the present application is shown. The multi-view cow face intelligent acquisition method comprises the following steps:

step 201: and acquiring video stream image data comprising the cow face and preprocessing the video stream image data.

In this embodiment, the electronic device (for example, the server/terminal device shown in fig. 1) on which the multi-view bovine face intelligent acquisition method operates may acquire video stream image data including bovine faces from the terminal device having a camera function through a wired connection manner or a wireless connection manner. It should be noted that the wireless connection means may include, but is not limited to, a 3G/4G/5G connection, a wifi connection, a bluetooth connection, a wimax connection, a zigbee connection, a UWB (u l t ra W i deband) connection, and other wireless connection means now known or developed in the future.

Further, preprocessing the video stream image data including the bovine face comprises:

and carrying out noise reduction, enhancement and normalization processing on the bovine face video stream data subjected to frame rate conversion.

Specifically, the frame rate and resolution of the video stream may be reduced or improved by performing image frame extraction (frame interpolation) and image filling on the obtained bovine face video data, for example, 1920 × 1080px60fps is converted into 640x480px30fps video, so as to obtain bovine face video stream data meeting the requirements; each frame of image in the acquired bovine face video stream image data can be converted into a certain image format, for example, the image is scaled to 320 size by taking the long edge as quasi-equal proportion, then preprocessing such as gaussian noise reduction, enhancement, normalization and the like is carried out on the image data, the data enhancement can be realized by adopting the modes of random transformation of image brightness, saturation and contrast, channel exchange, random horizontal turnover of images, motion blurring, median blurring, image quality compression, random rotation, random translation, multi-scale scaling, channel pixel dithering and the like, the normalization preprocessing can normalize the input bovine face image data to between [ -1,1], and high-quality and diverse bovine face image data can be obtained through the data preprocessing, so that the data can meet the standard of a pre-trained bovine face detection model, and the robustness and accuracy of the model are improved.

Step 202: the method comprises the steps of inputting the video stream image data after preprocessing into a trained cattle face detection model to obtain a cattle face detection result, wherein the cattle face detection result comprises cattle face images at various visual angles of multiple frames, and each frame of the cattle face image comprises a cattle face detection frame and corresponding cattle face key information.

Further, the training step of the cattle face detection model comprises:

the method for constructing the head network layer of the cattle face detection model according to the preset classification and/or regression tasks comprises the following steps: based on the basic convolution, the activation function and the bottleneck convolution, respectively constructing corresponding task branch networks to form the head network layer according to at least a preset cattle face central point thermodynamic diagram (center _ heat) classification task, a cattle face central point offset (center _ offset) regression task, a cattle face region frame width and height (wh) regression task, a cattle face key point thermodynamic diagram division (keyp i nt _ heat) class task and a cattle face key point offset (keypo i nt _ offset) regression task;

In the embodiment of the present invention, an overall structural schematic diagram of a constructed cattle face detection model is shown in fig. 3, where a cattle face detection model 300 mainly includes a backbone network layer 301, a neck network layer 302, and a head network layer 303, which are connected to each other; the backbone network layer of the bovine face detection model is constructed using ResNet50, and an image with 3 × 320 is input, and a bovine face feature map of 2048 × 10 is obtained after passing through the backbone network layer. The backbone network layer can extract the characteristics of the cattle face such as texture, color and scale in the cattle face image data and form a characteristic diagram for a subsequent network layer to use.

The neck network layer of the cattle face detection model is constructed by a convolution kernel (Conv) and a deconvolution kernel (Deconv), and is responsible for further extracting the characteristic information of the cattle face characteristic diagram output by the backbone network layer, and the extracted characteristic information can be composed and restored to the size of 1/4 of the original image, so that more accurate cattle face characteristic information can be extracted by the convolution and deconvolution up-sampling operation.

The above-mentioned header network layer can be composed of parallel 5 branches, all of which are composed of a basic convolution of 3*3, relu activation function and a convolution of 1*1, and its network structure is shown in fig. 4.

The cattle face detection model constructed in the embodiment of the invention is a single-stage detection network, can continuously detect cattle faces in videos at multiple visual angles (left, middle and right visual angles) from acquired video stream image data, and can obtain a plurality of frames of cattle face images at all visual angles, a corresponding cattle face detection frame, 5 cattle face key point positioning and other cattle face key information tasks; the cattle face key information extracted from the video stream image data through the cattle face detection model is beneficial to subsequent cattle face identification, and the prediction of the cattle face key points is based on thermodynamic diagrams, so that the cattle face detection model can more accurately locate the key points, and the algorithm is good in stability and high in efficiency.

It should be noted that the step of acquiring the cattle face data and performing data processing to obtain the cattle face data set includes manually collecting the cattle face data of different scenes and different cattle face angles, performing data labeling (such as key point labeling) and preprocessing such as data enhancement and normalization on the cattle face data, and randomly dividing the cattle face data set into a training set and a test set according to a certain proportion as required.

Step 203: and screening out multi-view cattle face images meeting preset quality standards from the cattle face images at all the view angles of the multi-frame according to the cattle face detection frame and corresponding key information of the cattle face and storing the multi-view cattle face images.

Further, according to the ox face detection frame and the key information of the corresponding ox face, the multi-view ox face image which accords with the preset quality standard is screened from the ox face images at each view angle of the multiframe, and the method comprises the following steps:

In the embodiment of the invention, whether the distance between a cow and the camera is proper or not can be judged according to the proportion of the cow detection frame in the cow image at each visual angle to the frame of the acquisition equipment, and a certain proportion threshold value is preset to screen the cow image meeting the requirement. For example, if the cow face detection frame is located near the center of the frame of the acquisition equipment at a certain viewing angle and the area ratio of the cow face detection frame is less than 50% of the frame, it is indicated that the cow is too far away from the camera at the moment, the corresponding image is discarded, and the camera of the camera is controlled to perform automatic and reasonable zooming operation, so that the cow face detection frame of the cow face image acquired at the corresponding viewing angle accounts for about 80% of the frame, the condition that a plurality of cow faces appear in the same frame of the acquisition equipment can be avoided, the clarity of the cow face image can be simultaneously detected, the acquired cow faces are prevented from being unclear and shielded, and the difficulty of a downstream cow face identification task is reduced; by repeating the process, a clear multi-view bovine face image set meeting the proportion threshold value can be obtained by screening the collected images.

Further, the step of screening out the multi-view cow face image meeting the preset quality standard from the cow face images at each view angle of the plurality of frames according to the cow face detection frame and the cow face key information further includes:

In the embodiment of the invention, the cow face plane relationship can be determined based on the cow face view angle types (left, middle and right faces) in the cow face key information and the cow face key point information, and the multi-view cow face image set is screened according to the cow face plane relationship; for example, for each frame of cattle face image of the multi-view cattle face image set, obtaining key information of the cattle face, and under the condition that the type of the cattle face view angle is a middle face, if three key points of two eyes and a nose tip of a cattle form an isosceles triangle, calculating an absolute value of an error of two base angles, if the absolute value of the error is less than 15 degrees, the middle face accords with an acquisition standard, so that the cattle face image under the middle face view angle is stored, otherwise, the cattle face image is directly discarded; under the condition that the type of the cow face visual angle is a left face, if an included angle formed by a vector from a nose tip to a left eye and a horizontal vector is smaller than 80 degrees and larger than 5 degrees and a key point of a right eye is invisible, the left face accords with an acquisition standard, so that the cow face image under the left face visual angle is stored, otherwise, the cow face image is directly discarded; under the condition that the type of the cow face visual angle is a right face, if an included angle formed by a vector from the nose tip to the right eye and the horizontal vector is smaller than 80 degrees and larger than 5 degrees and the key point of the left eye is invisible, the right face is considered to be in accordance with the acquisition standard, so that the cow face image under the right face visual angle is stored, and otherwise, the cow face image is directly discarded.

and further screening the multi-view cattle face image set according to the number of the cattle face detection frames of each frame of cattle face image and the positions of the cattle face detection frames to obtain the cattle face images under multiple views according with the quality standard.

In the embodiment of the invention, each frame of cattle face image in a multi-view cattle face image set detected by a cattle face detection model can be screened according to the number of the cattle face detection frames, if one frame of cattle face image corresponds to more than two cattle face detection frames, which indicates that the cattle face image contains more than two cattle faces at the same time, the corresponding cattle face image in the multi-view cattle face image set is discarded, and the cattle face image with only one cattle face detection frame is reserved; further, can also filter the ox face image of multi-view angle ox face image collection according to the position of ox face detection frame in the collection equipment frame, keep those ox face detection frame in the ox face image of collection equipment frame central point to put, get rid of those ox face detection frame too skew collection equipment frame central ox face image, and can filter the ox face image according to the definition, thereby can pass through above-mentioned screening process, the screening obtains the high quality ox face image under the multiple view angle that accords with preset quality standard from the ox face image at each visual angle of above-mentioned multiframe.

In summary, in the embodiments of the present invention, through acquiring video stream image data of a bovine face, performing preprocessing such as data enhancement and normalization, inputting a pre-trained bovine face detection model, detecting bovine face images of multiple frames at various viewing angles, and a bovine face detection frame and corresponding bovine face key information of each frame of bovine face image from the video stream image data through the bovine face detection model, and finally screening the bovine face images of the multiple frames at various viewing angles according to a preset quality standard based on the bovine face detection frame and the corresponding bovine face key information, a high-quality multi-viewing angle bovine face image meeting the standard is obtained for bovine face identification, so as to improve accuracy and robustness of bovine face identification. The cattle face detection model in the embodiment of the invention is a single-stage detection network, and can simultaneously perform tasks of multi-face (left, middle and right faces) cattle face detection and 5 cattle face key point positioning, wherein the prediction of the cattle face key points is based on thermodynamic diagrams, so that the target cattle face detection model can more accurately position the key points, and the algorithm has good stability and high efficiency.

The embodiment of the application can acquire and process related data based on an artificial intelligence technology. The artificial intelligence (Art I f I a l I nte l I gene, A I) is a theory, method, technology and application system which simulates, extends and expands human intelligence by using a digital computer or a machine controlled by the digital computer, senses the environment, acquires knowledge and obtains the best result by using the knowledge.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware associated with computer readable instructions, which can be stored in a computer readable storage medium, and when executed, the processes of the embodiments of the methods described above can be included. The storage medium may be a non-volatile storage medium such as a magnetic disk, an optical disk, a Read-only Memory (ROM), or a Random Access Memory (RAM).

It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless explicitly stated herein. Moreover, at least a portion of the steps in the flow chart of the figure may include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed alternately or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

With further reference to fig. 5, as an implementation of the method shown in fig. 2, the present application provides an embodiment of a multi-view intelligent cow face acquisition apparatus, where the embodiment of the apparatus corresponds to the embodiment of the method shown in fig. 2, and the apparatus may be specifically applied to various electronic devices.

As shown in fig. 5, the multi-view bovine face intelligent acquisition apparatus 500 according to the present embodiment includes:

an obtaining module 501, configured to obtain video stream image data including a cow face and perform preprocessing;

the detection module 502 is configured to input the preprocessed video stream image data into a trained cattle face detection model to obtain a cattle face detection result, where the cattle face detection result includes a plurality of frames of cattle face images at each viewing angle, a cattle face detection frame of each frame of cattle face image, and cattle face key information;

and the screening module 503 is configured to screen out and store multi-view bovine face images meeting preset quality standards from the bovine face images at the multiple viewing angles according to the bovine face detection frame and the bovine face key information.

The multi-view bovine face intelligent acquisition device in the embodiment of the invention can realize the function corresponding to the multi-view bovine face intelligent acquisition method and bring the same beneficial effects, and the repeated description is omitted here for avoiding the repetition.

In order to solve the technical problem, the embodiment of the application further provides a multi-view intelligent acquisition system for the cow face. Referring to fig. 6, fig. 6 is a schematic structural diagram of the multi-view bovine face intelligent acquisition system of the present embodiment. The multi-view bovine face intelligent acquisition system comprises a video stream image input and preprocessing module 601, a multi-view bovine face intelligent detection module 602, an intelligent feedback module 603 and a storage module 604, wherein the multi-view bovine face intelligent detection module 602 is used for detecting and screening out high-quality multi-view bovine face images from the video stream images, the intelligent feedback module 603 is used for carrying out feedback and prompt of acquisition operation to a video acquisition end or a user end according to the detected multi-view bovine face images, for example, the acquired image definition is fed back to the user, the acquisition end is prompted to adjust the acquired distance, angle, focal length and the like, and the storage module 604 is used for storing the acquired multi-view bovine face images locally and/or remotely.

In order to solve the technical problem, an embodiment of the present application further provides a computer device. Referring to fig. 7, fig. 7 is a block diagram of a basic structure of a computer device according to the present embodiment.

The computer device 7 comprises a memory 71, a processor 72, a network interface 73, which are communicatively connected to each other via a system bus. It is noted that only a computer device 7 having components 71-73 is shown, but it is to be understood that not all of the shown components are required to be implemented, and that more or fewer components may be implemented instead. AS will be understood by those skilled in the art, the computer device herein is a device capable of automatically performing numerical calculation and/or information processing according to a preset or stored instruction, and the hardware includes, but is not limited to, a microprocessor, an application specific integrated circuit (App I cat I on Spec I C I integrated C I rcu I, AS ic), a programmable Gate array (F I l D-programmable ab l Gate Ar ray, FPGA), a digital Processor (D I ta l S I gna l Processor, DSP), an embedded device, and the like.

The computer device can be a desktop computer, a notebook, a palm computer, a cloud server and other computing devices. The computer equipment can carry out man-machine interaction with a user through a keyboard, a mouse, a remote controller, a touch panel or voice control equipment and the like.

The memory 71 includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the storage 71 may be an internal storage unit of the computer device 7, such as a hard disk or a memory of the computer device 7. In other embodiments, the memory 71 may also be an external storage device of the computer device 7, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are provided on the computer device 7. Of course, the memory 71 may also comprise both an internal storage unit of the computer device 7 and an external storage device thereof. In this embodiment, the memory 71 is generally used for storing an operating system installed in the computer device 7 and various application software, such as computer readable instructions of the multi-view bovine face intelligent acquisition method. Further, the memory 71 may also be used to temporarily store various types of data that have been output or are to be output.

The processor 72 may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor, or other data Processing chip in some embodiments. The processor 72 is typically used to control the overall operation of the computer device 7. In this embodiment, the processor 72 is configured to execute computer readable instructions stored in the memory 71 or process data, for example, execute computer readable instructions of the multi-view bovine face intelligent acquisition method.

The network interface 73 may comprise a wireless network interface or a wired network interface, and the network interface 73 is generally used for establishing a communication connection between the computer device 7 and other electronic devices.

The present application further provides another embodiment, which is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, which can be executed by at least one processor, so as to cause the at least one processor to execute the steps of the multi-view bovine face intelligent acquisition method.

Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present application may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present application.

It is to be understood that the above-described embodiments are merely illustrative of some, but not restrictive, of the broad invention, and that the appended drawings illustrate preferred embodiments of the invention and do not limit the scope of the invention. This application is capable of embodiments in many different forms and is provided for the purpose of enabling a thorough understanding of the disclosure of the application. Although the present application has been described in detail with reference to the foregoing embodiments, it will be apparent to one skilled in the art that the present application may be practiced without modification or with equivalents of some of the features described in the foregoing embodiments. All equivalent structures made by using the contents of the specification and the drawings of the present application are directly or indirectly applied to other related technical fields and are within the protection scope of the present application.

Claims

1. A multi-view cow face intelligent acquisition method is characterized by comprising the following steps:

and screening out multi-view cattle face images meeting preset quality standards from the cattle face images at all the view angles of the multi-frame according to the cattle face detection frame and the key information of the cattle face, and storing the multi-view cattle face images.

2. The intelligent multi-view bovine face acquisition method according to claim 1, wherein the step of screening out the multi-view bovine face images meeting preset quality standards from the bovine face images at the various views of the plurality of frames according to the bovine face detection frame and the key information of the bovine face comprises:

3. The intelligent multi-view bovine face acquisition method according to claim 2, wherein the step of screening out multi-view bovine face images meeting preset quality standards from the bovine face images at various views of the plurality of frames according to the bovine face detection frame and the key information of bovine face comprises:

4. The method for intelligently acquiring multi-view bovine faces according to claim 3, wherein after the step of screening the multi-view bovine face image set according to the bovine face plane relationship, the method further comprises:

5. The method for intelligently acquiring multi-view bovine faces according to any one of claims 1-4, wherein the acquiring and preprocessing video stream image data including bovine faces comprises:

6. The multi-view intelligent bovine face acquisition method according to claim 5, wherein the training step of the bovine face detection model comprises:

the method for constructing the head network layer of the cattle face detection model according to the preset classification and/or regression task comprises the following steps: based on basic convolution, an activation function and bottleneck convolution, respectively constructing corresponding task branch networks to form the head network layer according to at least a preset thermodynamic diagram classification task of a cattle face central point, an offset regression task of the cattle face central point, a width and height regression task of a cattle face detection frame, a thermodynamic diagram classification task of the cattle face key points and an offset regression task of the cattle face key points;

7. The utility model provides a multi-view ox face intelligent acquisition device which characterized in that includes:

the acquisition module is used for acquiring and preprocessing video stream image data comprising the cow face;

8. The utility model provides a multi-view ox face intelligent acquisition system, a serial communication port, including video stream image input and preprocessing module, multi-view ox face intelligent detection module, intelligent feedback module and storage module, wherein multi-view ox face intelligent detection module is arranged in detecting from the video stream image and selects high-quality multi-view ox face image with screening, intelligent feedback module is used for carrying out the feedback and the suggestion of gathering the operation to video acquisition end and/or user side according to the multi-view ox face image that detects out, storage module is used for the multi-view ox face image of local and/or remote storage collection.

9. A computer device comprising a memory having computer readable instructions stored therein and a processor that when executed performs the steps of the multi-perspective bovine face intelligent acquisition method of any one of claims 1 to 6.

10. A computer readable storage medium, wherein the computer readable storage medium stores thereon computer readable instructions, which when executed by a processor, implement the steps of the multi-view bovine face intelligent acquisition method according to any one of claims 1 to 6.