Dairy cow posture index monitoring method and system

Abstract

The invention provides a dairy cow posture index monitoring method and system, and relates to the technical field of farming and pastoral breeding, and the method comprises the steps: obtaining monitoring data of a target dairy cow, the monitoring data comprising a preset number of plane images and a stereoscopic video of the target dairy cow in a period of time; preprocessing the plane image, and selecting a target plane image from all the preprocessed plane images; carrying out identity verification on the target dairy cow by using the target plane image; after the target dairy cow passes the identity verification, selecting a target model corresponding to the current monitoring index from a plurality of preset models; the models are in one-to-one correspondence with the monitoring indexes; and based on the target model, according to the stereoscopic video, determining value data of the target dairy cow about the current monitoring index. According to the method, the dairy cow posture index evaluation efficiency and accuracy are improved, the workload of evaluation personnel is reduced, and the threshold of the evaluation personnel is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of agriculture and animal husbandry, and in particular to a method and system for monitoring body indicators of dairy cows. Background Art

[0002] With the development of large-scale and mechanized dairy farming, the market's requirements for the body shape, appearance and behavior of dairy cows are also increasing. Monitoring the body shape, appearance and behavior of dairy cows not only helps to understand the health level and growth and development of dairy cows, but is also directly related to the genetic potential and breeding value of dairy cows.

[0003] Currently, professional evaluators often visually observe and judge dairy cows, and score them based on their overall body shape, proportions of various parts, body structure, walking posture, lying still, skin and coat color, etc. This method requires high professional knowledge and experience of the staff, and the evaluation efficiency and accuracy are low.

[0004] Based on this, how to lower the threshold and workload of evaluators and improve evaluation efficiency and accuracy has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] In view of this, in order to solve the above technical problems, the present invention provides a method and system for monitoring body posture indicators of dairy cows.

[0006] The present invention adopts the following technical solution: In a first aspect, the present invention provides a method for monitoring body indicators of dairy cows, comprising: Acquire monitoring data of a target cow, wherein the monitoring data includes a preset number of planar images and a stereoscopic video of the target cow within a period of time; Preprocessing the plane images, and selecting a target plane image from all the plane images after preprocessing; Authentication of the target cow using the target plane image; After the target cow passes identity authentication, a target model corresponding to the current monitoring index is selected from a plurality of preset models; the model corresponds to the monitoring index one by one; Based on the target model and according to the stereoscopic video, the value data of the target cow with respect to the current monitoring index is determined.

[0007] Optionally, obtain monitoring data of the target cow, including: Acquire a preset number of planar images of the target cow by using a binocular visible light camera; A stereoscopic video of the target cow within a period of time is obtained through an intelligent wide-spectrum sensor.

[0008] Optionally, preprocessing the planar image specifically includes: The planar image is cleaned, enhanced and grayed.

[0009] Optionally, using the target plane image to authenticate the target cow specifically includes: Extracting texture features from the target plane image to obtain target texture features; Determine whether there is a texture feature corresponding to the target texture feature among all texture features stored in a preset texture feature library; If there is a texture feature corresponding to the target texture feature among all the texture features stored in the preset texture feature library, determining the identity code associated with the corresponding texture feature as the first identity code of the target cow; It is determined that the target cow passes identity authentication.

[0010] Optionally, the monitoring data also includes a second identity code read by an RFID reader; Before determining that the target cow has passed identity authentication, the cow body index monitoring method of the present invention further includes: Determining whether the first identity code is consistent with the second identity code; If the first identity code is consistent with the second identity code, it is determined that the target cow has passed the identity authentication.

[0011] Optionally, the monitoring indicators include body shape and appearance indicators and behavior and posture indicators.

[0012] Optionally, the body shape and appearance indicators specifically include cow side measurement parameters and cow back measurement parameters; The cow side measurement parameters at least include udder area, udder height, body height, body depth, shoulder height, hip height and rump height; The cow back measurement parameters at least include body width, hip width and rump width.

[0013] Optionally, the behavioral posture indicators include at least nodding, hunching the back, and swaying gait.

[0014] Optionally, the method for monitoring the body indicators of dairy cows of the present invention further includes: After obtaining the value data of each monitoring indicator of the target dairy cow, using all the value data to generate a monitoring report; The monitoring report is sent to a preset smart terminal so that a user can view the monitoring report through the preset smart terminal.

[0015] In a second aspect, the present invention provides a system for monitoring the body indicators of dairy cows, which is applied to the method for monitoring the body indicators of dairy cows as described above, and the monitoring system includes: a data acquisition terminal, a local controller and a cloud server; The data acquisition terminal is used to obtain monitoring data of the target cow and send the monitoring data to the local controller, wherein the monitoring data includes a preset number of plane images and a stereoscopic video of the target cow within a period of time; The local controller is used to pre-process the plane images, select a target plane image from all the plane images after pre-processing, and send the target plane image to the cloud server; The cloud server is used to authenticate the target cow using the target plane image and return a verification result to the local controller; The local controller is also used to pre-process the stereoscopic video after determining that the target cow has passed the identity authentication according to the verification result, and send the pre-processed stereoscopic video to the cloud server; The cloud server is also used to select a target model corresponding to the current monitoring indicator from a plurality of preset models, and based on the target model and the preprocessed stereo video, determine the value data of the target cow regarding the current monitoring indicator, and the model corresponds one-to-one to the monitoring indicator.

[0016] The present invention adopts the above technical scheme, a method for monitoring the body indicators of dairy cows, comprising: obtaining monitoring data of a target dairy cow, the monitoring data comprising a preset number of plane images and a stereoscopic video of the target dairy cow within a period of time; preprocessing the plane images, and selecting a target plane image from all the plane images after preprocessing; using the target plane image to authenticate the identity of the target dairy cow; after the target dairy cow passes the identity authentication, selecting a target model corresponding to the current monitoring indicator from a plurality of preset models; the model corresponds to the monitoring indicator one by one; based on the target model, according to the stereoscopic video, determining the value data of the target dairy cow regarding the current monitoring indicator.

[0017] Based on this, the present invention can automatically obtain monitoring data of dairy cows, and automatically use the monitoring data to calculate the value data of the dairy cows regarding the current monitoring indicators, thereby realizing automated monitoring. The present invention improves the evaluation efficiency and accuracy, reduces the workload of evaluators, and the present invention does not have too high requirements on the professional knowledge and experience of the staff, thereby lowering the threshold for evaluators. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is a flow chart of a method for monitoring the body indicators of dairy cows provided by an embodiment of the present invention; Figure 2 The present invention is a schematic diagram of the structure of a dairy cow body index monitoring system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0021] Example Figure 1 FIG. 1 is a flow chart of a method for monitoring the body indicators of dairy cows provided by an embodiment of the present invention. Figure 1 As shown, this process includes: Step 101: Acquire monitoring data of a target cow, where the monitoring data includes a preset number of planar images and a stereoscopic video of the target cow within a period of time.

[0022] In the embodiment of the present invention, obtaining the monitoring data of the target dairy cow may specifically include: A preset number of planar images of the target cow are obtained through a binocular visible light camera, and a stereoscopic video of the target cow over a period of time is obtained through an intelligent wide-spectrum sensor.

[0023] Specifically, the binocular visible light camera and the intelligent wide spectrum sensor are fixed at corresponding positions respectively, and the distance between them and the center of the preset shooting area, the angle relative to the ground and the vertical height all meet the relevant requirements, so that after the cow enters the preset shooting area, the binocular visible light camera and the intelligent wide spectrum sensor can respectively capture the planar image and stereo video of the cow.

[0024] The binocular visible light camera and the intelligent wide spectrum sensor can be installed on the collection channel. The binocular visible light camera can take continuous photos of the cow after determining that the cow has entered the preset shooting area, and obtain a preset number of flat images. The intelligent wide spectrum sensor can capture a video of the cow walking when the cow walks through the collection channel, and obtain a stereoscopic video of the cow over a period of time. Both the binocular visible light camera and the intelligent wide spectrum sensor have the characteristics of high resolution and low distortion, ensuring that clear and comprehensive images of the cow's body can be collected.

[0025] Step 102: preprocessing the plane images, and selecting a target plane image from all the plane images after preprocessing.

[0026] Specifically, for each plane image, image preprocessing technology is used to clean, enhance and grayscale the image, so as to enhance the monitorability of the information about the body characteristics of the cow. Among them, image cleaning can remove abnormal images; image enhancement can adjust the visual effect to facilitate segmentation and detection; grayscale processing can convert color images into grayscale images to reduce the amount of data and increase processing speed.

[0027] After all the plane images are preprocessed, an optimal image can be selected from the currently remaining plane images for identity verification of the target cow.

[0028] Step 103: Use the target plane image to authenticate the target cow.

[0029] In the embodiment of the present invention, using the target plane image to authenticate the target cow may specifically include: (1031) Texture features are extracted from the target plane image to obtain target texture features.

[0030] Specifically, the texture features in the plane image can be automatically learned and extracted based on the deep learning algorithm. The specific learning and extraction are existing technologies and will not be described in detail here.

[0031] (1032) It is determined whether there is a texture feature corresponding to the target texture feature among all the texture features stored in the preset texture feature library.

[0032] Specifically, the texture features of all cows can be stored in advance to obtain a preset texture feature library. The preset texture feature library includes the texture features of all cows of the user and their associated identity codes (identity codes of cows).

[0033] (1033) If there is a texture feature corresponding to the target texture feature among all the texture features stored in the preset texture feature library, then the identity code associated with the corresponding texture feature is determined as the first identity code of the target cow.

[0034] (1034) Determine if the target cow has passed identity verification.

[0035] In the embodiment of the present invention, in order to improve the accuracy of the cow identity authentication result, the monitoring data may further include a second identity code read by the RFID reader.

[0036] Specifically, the present invention can deploy an individual cow IoT system to further verify the cow's identity through RFID (Radio Frequency Identification) technology. In practical applications, each cow will wear an ear tag (RFID tag), each ear tag has a unique number, and the second identity code stored in the ear tag is read by an RFID reader. The RFID reader position setting principle is the same or similar to the setting principle of the aforementioned binocular visible light camera and intelligent wide spectrum sensor.

[0037] Before determining that the target cow has passed identity authentication, the cow body index monitoring method of the present invention may further include: It is determined whether the first identity code is consistent with the second identity code. If the first identity code is consistent with the second identity code, it is determined that the target cow has passed the identity authentication. In this way, by performing double identity authentication on the cow, the accuracy of the cow's identity authentication result is improved.

[0038] Step 104: After the target cow passes identity authentication, a target model corresponding to the current monitoring index is selected from a plurality of preset models; the model corresponds to the monitoring index one by one.

[0039] Specifically, the present invention pre-trains a corresponding model for each monitoring indicator, and the model corresponding to any monitoring indicator is an image segmentation model or a key point detection model; when the monitoring indicator is a range value indicator such as breast area, the corresponding model is an image segmentation model; when the monitoring indicator is a specific value indicator such as body depth, the corresponding model is a key point detection model.

[0040] The image segmentation model extracts and segments the cow image based on the UNet technology of CNN (Convolutional Neural Network), and calculates the average value of the segmented image based on data calculation methods such as thresholds, edges, regions, and deep learning to obtain the value of the monitoring indicator. For example, when the current monitoring indicator is the udder area, the image segmentation model calculates multiple udder areas using stereoscopic video, and then calculates the average value of each udder area and outputs the average value (the value of the monitoring indicator). It should be noted that the image segmentation model extracts and segments the cow image based on the UNet technology of CNN, and calculates the average value of the segmented image based on data calculation methods such as thresholds, edges, regions, and deep learning. The technologies used in this process are all existing technologies and will not be repeated here.

[0041] The key point detection model performs key point recognition and detection on the features of the cow image based on the deep learning YOLO (You Only Look Once) v5, and calculates the average value by combining regression method or curve fitting method to obtain the value of the monitoring indicator. For example, when the current monitoring indicator is nodding, when the key point detection model uses stereo video to calculate multiple nodding depths, the key point detection model calculates the average value of each nodding depth and outputs the average value (the value of the monitoring indicator). It should be noted that the key point detection model performs key point recognition and detection on the features of the cow image based on the deep learning YOLO (You Only Look Once) v5, and calculates the average value by combining regression method or curve fitting method to obtain the value of the monitoring indicator. The technologies used in this process are all existing technologies and will not be repeated here.

[0042] In addition, the training method of the model (image segmentation model and key point detection model) used in the present invention includes: data collection-data annotation-data preprocessing-model construction (selecting a model and adjusting parameters)-training model (repeated multiple times until the effect is achieved)-testing model-adjusting parameters until the optimal model is trained. It should be noted that the training method of the model (image segmentation model and key point detection model) used in the present invention is also a prior art and will not be described in detail here.

[0043] Step 105: Based on the target model and the stereoscopic video, the value data of the target cow regarding the current monitoring index is determined.

[0044] Specifically, after preprocessing each frame of the stereoscopic video, each preprocessed image is input into the target model to obtain the value data of the target cow on the current monitoring index output by the target model.

[0045] In the embodiment of the present invention, the monitoring indicators include body shape and appearance indicators and behavior and posture indicators.

[0046] Among them, the body shape and appearance indicators specifically include the cow's side measurement parameters and the cow's back measurement parameters; the cow's side measurement parameters at least include udder area, udder height, body height, body depth, shoulder height, hip height and buttocks height; the cow's back measurement parameters at least include body width, hip width and buttocks width.

[0047] Behavioral posture indicators include at least nodding, arching the back, and swaying gait.

[0048] In an embodiment of the present invention, the method for monitoring the body indicators of dairy cows of the present invention further includes: (1) After obtaining the value data of each monitoring indicator of the target dairy cow, use all the value data to generate a monitoring report.

[0049] (2) The monitoring report is sent to a preset smart terminal so that the user can view the monitoring report through the preset smart terminal.

[0050] Specifically, the preset smart terminal can be a smart phone or a computer. Farmers can check the monitoring report at any time through the preset smart terminal to understand the physical condition of the cows. Farmers can also predict the health trend of the cows, the growth and development trend of the cows, and the future production of the cows based on the body shape and behavior posture parameters obtained through long-term monitoring.

[0051] The embodiment of the present invention adopts the above-mentioned technical solution, which can automatically obtain the monitoring data of the dairy cow, and automatically use the monitoring data to calculate the value data of the dairy cow regarding the current monitoring index, thereby realizing automated monitoring. The present invention improves the evaluation efficiency and accuracy, reduces the workload of the evaluator, and the present invention does not have too high requirements on the professional knowledge and experience of the staff, thereby lowering the threshold for the evaluator.

[0052] Furthermore, the present invention collects monitoring data of dairy cows through binocular visible light cameras and intelligent wide-spectrum sensors, realizes non-contact monitoring, improves the real-time and reliability of data collection, and overcomes the problem that traditional manual measurement easily interferes with dairy cows and causes stress reactions in dairy cows. Furthermore, the present invention integrates advanced technologies such as computer vision and artificial intelligence, and can automatically learn and extract parameters such as udder area, body depth, and body height of dairy cows, realizing intelligent evaluation.

[0053] Based on a general inventive concept, the present invention also provides a system for monitoring the body indicators of dairy cows, which is applied to the method for monitoring the body indicators of dairy cows as described above.

[0054] Figure 2 Schematic diagram of a cow body index monitoring system provided by an embodiment of the present invention. Figure 2As shown, the system includes: a data acquisition terminal 21, a local controller 22 and a cloud server 23.

[0055] The data acquisition terminal 21 is used to obtain the monitoring data of the target cow and send the monitoring data to the local controller. The monitoring data includes a preset number of plane images and a stereoscopic video of the target cow within a period of time. In practical applications, the data acquisition terminal 21 can specifically include the aforementioned binocular visible light camera, intelligent wide spectrum sensor and RFID reader / writer, and the data acquisition terminal 21 can be set in the channel at the entrance or exit of the milking parlor. Specifically, the data acquisition terminal 21 can be installed by modifying the original channel at the entrance or exit of the milking parlor.

[0056] The local controller 22 is used to pre-process the plane images, select a target plane image from all the plane images after the pre-processing, and send the target plane image to the cloud server.

[0057] The cloud server 23 is used to authenticate the target cow using the target plane image and return the authentication result to the local controller.

[0058] The local controller 22 is also used to pre-process the stereoscopic video after determining that the target cow has passed the identity authentication according to the verification result, and send the pre-processed stereoscopic video to the cloud server.

[0059] The cloud server 23 is also used to select a target model corresponding to the current monitoring index from a plurality of preset models, and based on the target model and the preprocessed stereo video, determine the value data of the target cow regarding the current monitoring index, so that the model corresponds to the monitoring index one by one.

[0060] It should be noted that the cow posture index monitoring system of this embodiment and the aforementioned cow posture index monitoring method are based on a general inventive concept and have the same or corresponding execution process. Therefore, the specific implementation details can be found in the aforementioned embodiments and will not be repeated here.

[0061] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0062] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0063] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present invention belong.

[0064] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0065] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0066] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0067] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0068] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0069] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for monitoring the body indicators of dairy cows, characterized in that: include: Acquire monitoring data of a target cow, wherein the monitoring data includes a preset number of planar images and a stereoscopic video of the target cow within a period of time; Preprocessing the plane images, and selecting a target plane image from all the plane images after preprocessing; Authentication of the target cow using the target plane image; After the target cow passes identity authentication, a target model corresponding to the current monitoring index is selected from a plurality of preset models; the model corresponds to the monitoring index one by one; Based on the target model and according to the stereoscopic video, the value data of the target cow with respect to the current monitoring index is determined.

2. The method for monitoring the body indicators of dairy cows according to claim 1, characterized in that: Obtain monitoring data for target dairy cows, including: Acquire a preset number of planar images of the target cow by using a binocular visible light camera; A stereoscopic video of the target cow within a period of time is obtained through an intelligent wide-spectrum sensor.

3. The method for monitoring the body indicators of dairy cows according to claim 1, characterized in that: Preprocessing the planar image specifically includes: The planar image is cleaned, enhanced and grayed.

4. The method for monitoring the body indicators of dairy cows according to claim 1, characterized in that: Using the target plane image to authenticate the target cow specifically includes: Extracting texture features from the target plane image to obtain target texture features; Determine whether there is a texture feature corresponding to the target texture feature among all texture features stored in a preset texture feature library; If there is a texture feature corresponding to the target texture feature among all the texture features stored in the preset texture feature library, determining the identity code associated with the corresponding texture feature as the first identity code of the target cow; It is determined that the target cow passes identity authentication.

5. The method for monitoring the body indicators of dairy cows according to claim 4, characterized in that: The monitoring data also includes a second identity code read by the RFID reader; Before determining that the target cow has passed the identity authentication, it also includes: Determining whether the first identity code is consistent with the second identity code; If the first identity code is consistent with the second identity code, it is determined that the target cow has passed the identity authentication.

6. The method for monitoring the body indicators of dairy cows according to claim 1, characterized in that: The monitoring indicators include body shape and appearance indicators and behavior and posture indicators.

7. The method for monitoring the body indicators of dairy cows according to claim 6, characterized in that: The body shape and appearance indicators specifically include cow side measurement parameters and cow back measurement parameters; The cow side measurement parameters at least include udder area, udder height, body height, body depth, shoulder height, hip height and rump height; The cow back measurement parameters at least include body width, hip width and rump width.

8. The method for monitoring the body indicators of dairy cows according to claim 6, characterized in that: The behavioral posture indicators at least include nodding, hunching back and swaying gait.

9. The method for monitoring the body indicators of dairy cows according to claim 1, characterized in that: Also includes: After obtaining the value data of each monitoring indicator of the target dairy cow, using all the value data to generate a monitoring report; The monitoring report is sent to a preset smart terminal so that a user can view the monitoring report through the preset smart terminal.

10. A system for monitoring the body indicators of dairy cows, applied to the method for monitoring the body indicators of dairy cows as claimed in any one of claims 1 to 9, characterized in that: include: Data acquisition terminal, local controller and cloud server; The data acquisition terminal is used to obtain monitoring data of the target cow and send the monitoring data to the local controller, wherein the monitoring data includes a preset number of plane images and a stereoscopic video of the target cow within a period of time; The local controller is used to pre-process the plane images, select a target plane image from all the plane images after pre-processing, and send the target plane image to the cloud server; The cloud server is used to authenticate the target cow using the target plane image and return a verification result to the local controller; The local controller is also used to pre-process the stereoscopic video after determining that the target cow has passed the identity authentication according to the verification result, and send the pre-processed stereoscopic video to the cloud server; The cloud server is also used to select a target model corresponding to the current monitoring indicator from a plurality of preset models, and based on the target model and the preprocessed stereoscopic video, determine the value data of the target cow regarding the current monitoring indicator, and the model corresponds one-to-one to the monitoring indicator.

Images