Veterinary backfat acquisition method and apparatus

Abstract

A veterinary backfat acquisition method and an apparatus. The method comprises: on the basis of a coordinate positioning numerical value and a numerical value of the distance from a depth camera to the back of an acquisition subject, generating a mechanical arm control instruction, sending the mechanical arm control instruction to a multi-axis mechanical arm, and, on the basis of the mechanical arm control instruction, the multi-axis mechanical arm moving according to a preset trajectory to a couplant container to dip a couplant; after dipping the couplant has been done, controlling a backfat B-ultrasound probe mounted at the top of the multi-axis mechanical arm to arrive at a measurement point on the acquisition subject and attaching same to the skin of the acquisition subject; and after the backfat B-ultrasound probe has arrived at the backfat measurement point, repeatedly acquiring backfat B-ultrasound data according to a preset number of times, and transmitting to a master control cabinet the acquired B-ultrasound data in the form of a video. The present invention can achieve automated acquisition, thus greatly reducing the labor cost; the present invention can accurately position backfat measurement positions of pigs; in addition, the present invention reduces links in which humans are in contact with livestock, thus improving the biosafety level and reducing the risk of infectious diseases brought by humans.

Description

Animal backfat collection method and device

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on November 2, 2023, with application number 202311452550.2 and application name “A method and device for collecting back fat for animals”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The invention belongs to the technical field of backfat collection, and in particular relates to a backfat collection method and device for animals. Background Art

[0003] At present, the backfat of pigs is mainly measured manually using veterinary B-ultrasound instruments. During the measurement work, specific cages need to be placed in a fixed place, and then the pigs that need to have their backfat measured are driven into the specific cages one by one. The staff then finds the backfat measurement point of the pig based on their experience, and then conducts B-ultrasound monitoring. If the staff lacks experience, it may take several minutes or even more than ten minutes to find the point. In the process of searching for the backfat measurement point, it is likely to cause an overreaction in the pig. After finding the backfat measurement point, the staff uses a handheld B-ultrasound instrument to perform B-ultrasound testing on the pig, and then uses the handheld instrument to measure the backfat value, and manually records it after the operation is completed.

[0004] During this work process, pigs are easily subjected to stress reactions due to being away from their original environment and being driven away, and they become more irritable during the measurement work, so the backfat measurement work is sometimes very difficult. In addition, the backfat measurement position is located manually based on experience, but due to the different experience of the staff, the measurement point positioning is often incorrect, which often leads to large errors in the measurement data.

[0005] Summary of the Invention

[0006] To this end, the present invention provides a method and device for collecting backfat for animals, which solves the problems of traditional veterinary backfat collection, such as long time, high difficulty and low data accuracy.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a method for collecting back fat for animals, comprising:

[0008] S1. While the subject is eating, an electronic ear tag reader / writer identifies the electronic ear tag on the subject's ear and obtains the ID of the electronic ear tag of the subject;

[0009] S2. After receiving the number of the electronic ear tag of the object to be collected, the main control cabinet starts the depth camera to collect images of the object to be collected, and uses the backfat measurement point recognition model to locate the coordinates of the backfat measurement point on the collected image;

[0010] S3. Generate a robotic arm control instruction based on the coordinate positioning value and the distance value from the depth camera to the back of the acquisition object, and send the robotic arm control instruction to the multi-axis robotic arm. The multi-axis robotic arm moves to the coupling agent container along a preset trajectory according to the robotic arm control instruction to dip the coupling agent;

[0011] S4. After the coupling agent is applied, the multi-axis robotic arm is controlled to move at a first speed to a given coordinate position until a distance sensor on the multi-axis robotic arm detects that a backfat B-ultrasound probe mounted on the top of the multi-axis robotic arm reaches a point on the back of the subject where backfat measurement is required. Then, the multi-axis robotic arm is controlled to move at a second speed until it contacts the subject's skin.

[0012] S5. When the backfat B-ultrasound probe reaches the backfat measurement point, it repeatedly collects backfat B-ultrasound data according to a set number of times, and transmits the collected B-ultrasound data to the main control cabinet in the form of video.

[0013] As a preferred embodiment of the veterinary backfat collection method, in step S2, the training process of the backfat measurement point recognition model includes:

[0014] Acquiring back image data of the acquisition object;

[0015] Annotating the image data of the back of the collected object;

[0016] Load the YoLov4-tiny model for model training;

[0017] Use the test set to test the trained model;

[0018] Evaluate the test results and end the training if they meet the set requirements.

[0019] As a preferred embodiment of the veterinary back fat collection method, in step S4, the multi-axis robotic arm is controlled to move at a second speed until it fits the skin of the collection object, and then the multi-axis robotic arm is continued to be controlled to move so that the skin of the collection object is pressed down a preset distance.

[0020] As a preferred embodiment of the veterinary backfat collection method, step S4 also includes determining whether the position of the collection object has moved. If the position of the collection object has moved, the position of the multi-axis robotic arm is adjusted according to the movement of the collection object so that the backfat B-ultrasound probe continues to be fully in contact with the back of the collection object.

[0021] As a preferred embodiment of the veterinary backfat collection method, in step S5, when the backfat B-ultrasound data is collected a set number of times, the backfat B-ultrasound probe is controlled to stop data collection; if the collection object changes the feeding position, the backfat measurement and collection of the collection object is performed again;

[0022] In step S5, the collected B-ultrasound data is stored locally, and a chart including the measurement date, ear tag number, and B-ultrasound image content is generated.

[0023] The present invention also provides a backfat collection system for animals, comprising:

[0024] An ear tag number acquisition module is used to identify the electronic ear tag on the ear of the collection subject through the electronic ear tag reader and writer during the collection subject's eating process, and obtain the number of the electronic ear tag of the collection subject;

[0025] The image acquisition module is used to start the depth camera to acquire an image of the acquisition object after the main control cabinet receives the number of the electronic ear tag of the acquisition object;

[0026] A coordinate positioning module, used for positioning the backfat measurement points on the captured image using a backfat measurement point recognition model;

[0027] A robot arm control instruction generation module is used to generate a robot arm control instruction according to the coordinate positioning value and the distance value from the depth camera to the back of the acquisition object;

[0028] a coupling agent dipping control module, configured to send the robotic arm control instruction to the multi-axis robotic arm, so that the multi-axis robotic arm moves to the coupling agent container along a preset trajectory to dip the coupling agent according to the robotic arm control instruction;

[0029] a robotic arm motion control module, configured to, after the coupling agent is applied, control the multi-axis robotic arm to move to a given coordinate position at a first speed until a distance sensor on the multi-axis robotic arm detects that a backfat B-ultrasound probe mounted on the top of the multi-axis robotic arm has reached a point on the back of the subject where backfat measurement is required, and then control the multi-axis robotic arm to move at a second speed until it contacts the subject's skin;

[0030] The backfat B-ultrasound data acquisition module is used to repeatedly acquire backfat B-ultrasound data according to a set number of times after the backfat B-ultrasound probe reaches the backfat measurement point, and transmit the acquired B-ultrasound data to the main control cabinet in the form of video.

[0031] As a preferred solution of the veterinary backfat collection system, a model training module is further included for training the backfat measurement point recognition model; the model training module includes:

[0032] An image loading submodule, configured to obtain image data of the back of the subject;

[0033] An image data annotation submodule, configured to annotate the image data of the back of the captured object;

[0034] Model loading submodule, used to load the YoLov4-tiny model for model training;

[0035] The model testing submodule is used to test the trained model using the test set;

[0036] The test evaluation submodule is used to evaluate the test results and end the training if the set requirements are met.

[0037] As a preferred solution for the veterinary back fat collection system, the robotic arm motion control module is also used to control the multi-axis robotic arm to move at a second speed until it is in contact with the skin of the collection object, and then continue to control the movement of the multi-axis robotic arm to press the skin of the collection object down a preset distance.

[0038] As a preferred solution for the veterinary back fat collection system, it also includes a robotic arm position adjustment module, which is used to determine whether the position of the collection object has moved. If the position of the collection object has moved, the position of the multi-axis robotic arm is adjusted according to the movement of the collection object so that the back fat B-ultrasound probe continues to maintain complete fit with the back of the collection object.

[0039] As a preferred solution of the veterinary backfat collection system, in the backfat B-ultrasound data collection module, when the backfat B-ultrasound data is collected a set number of times, the backfat B-ultrasound probe is controlled to stop data collection; if the collection object changes the feeding position, the backfat measurement and collection of the collection object is performed again;

[0040] It also includes a data storage module for locally storing the collected B-ultrasound data and generating a chart containing the measurement date, ear tag number, and B-ultrasound image content.

[0041] The beneficial effects of the present invention are as follows: during the eating process of the collection object, the electronic ear tag on the ear of the collection object is identified by the electronic ear tag reader, and the number of the electronic ear tag of the collection object is obtained; after the main control cabinet receives the number of the electronic ear tag of the collection object, the depth camera is started to collect images of the collection object, and the backfat measurement point recognition model is used to coordinate the backfat measurement point on the collected image; according to the coordinate positioning value and the distance value from the depth camera to the back of the collection object, a robot arm control instruction is generated, and the robot arm control instruction is sent to the multi-axis robot arm, and the multi-axis robot arm is controlled according to the mechanical The arm control instruction moves to the coupling agent container according to a preset trajectory to dip the coupling agent; when the coupling agent is dipped, the multi-axis robotic arm is controlled to move to a given coordinate position at a first speed until the distance sensor on the multi-axis robotic arm detects that the backfat B-ultrasound probe installed on the top of the multi-axis robotic arm reaches the point on the back of the collection object where backfat measurement is required, and then the multi-axis robotic arm is controlled to move at a second speed until it fits the skin of the collection object; when the backfat B-ultrasound probe reaches the backfat measurement point, the backfat B-ultrasound data is repeatedly collected according to the set number of times, and the collected B-ultrasound data is transmitted to the main control cabinet in the form of video. The present invention can realize automated collection, greatly saving labor costs; can accurately locate the backfat measurement position of pigs; can perform image collection during the measurement process and save the measurement images during B-ultrasound measurement, which is convenient for staff at all levels to review the measurement data; automatically perform backfat measurement during the feeding stage of the collection object, which can greatly reduce the stress effect on the collection object during the measurement work; reduce the contact between humans and animals, improve the biosafety level, and reduce the risk of infectious diseases brought by humans. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are merely exemplary, and those skilled in the art can, without inventive effort, derive other implementation drawings based on the provided drawings.

[0043] The structures, proportions, sizes, etc. illustrated in this specification are intended solely to complement the contents disclosed herein and to facilitate understanding and reading by persons skilled in the art. They are not intended to limit the conditions under which the present invention may be implemented and therefore have no substantive technical significance. Any structural modifications, changes in proportions, or adjustments in sizes, without affecting the efficacy and objectives of the present invention, shall remain within the scope of the technical contents disclosed herein.

[0044] FIG1 is a flow chart of a method for collecting backfat for animals according to an embodiment of the present invention;

[0045] FIG2 is a schematic diagram of a model training process for a veterinary backfat collection method according to an embodiment of the present invention;

[0046] FIG3 is a schematic diagram of hardware deployment of a veterinary backfat collection method according to an embodiment of the present invention;

[0047] FIG4 is a schematic diagram of the architecture of a veterinary backfat collection system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following describes the implementation of the present invention using specific embodiments. Those skilled in the art will readily understand the other advantages and benefits of the present invention from the disclosure herein. Obviously, the embodiments described are only a portion of the present invention, not all of it. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are intended to fall within the scope of protection of the present invention.

[0049] Example 1

[0050] 1 and 2 , embodiment 1 of the present invention provides a method for collecting backfat for animals, comprising the following steps:

[0051] S1. While the subject is eating, an electronic ear tag reader / writer identifies the electronic ear tag on the subject's ear and obtains the ID of the electronic ear tag of the subject;

[0052] S2. After receiving the number of the electronic ear tag of the object to be collected, the main control cabinet starts the depth camera to collect images of the object to be collected, and uses the backfat measurement point recognition model to locate the coordinates of the backfat measurement point on the collected image;

[0053] S3. Generate a robotic arm control instruction based on the coordinate positioning value and the distance value from the depth camera to the back of the acquisition object, and send the robotic arm control instruction to the multi-axis robotic arm. The multi-axis robotic arm moves to the coupling agent container along a preset trajectory according to the robotic arm control instruction to dip the coupling agent;

[0054] S4. After the coupling agent is applied, the multi-axis robotic arm is controlled to move at a first speed to a given coordinate position until a distance sensor on the multi-axis robotic arm detects that a backfat B-ultrasound probe mounted on the top of the multi-axis robotic arm reaches a point on the back of the subject where backfat measurement is required. Then, the multi-axis robotic arm is controlled to move at a second speed until it contacts the subject's skin.

[0055] S5. When the backfat B-ultrasound probe reaches the backfat measurement point, it repeatedly collects backfat B-ultrasound data according to a set number of times, and transmits the collected B-ultrasound data to the main control cabinet in the form of video.

[0056] In this embodiment, in step S1, taking the pig as an example, when the pig's head reaches the hopper to eat, the electronic ear tag reader recognizes the electronic ear tag on the pig's ear and then transmits the electronic ear tag number to the main control cabinet.

[0057] In step S2, when the main control cabinet receives the electronic ear tag number, it starts to start the depth camera to capture images of the pig body, and locates the coordinates of the backfat measurement point that needs to be measured according to the backfat measurement point recognition model, and obtains the distance value from the camera to the pig's back, and then transmits the positioning coordinates and distance value to the main control cabinet; the positioning method of the backfat measurement point is model training, which has very high accuracy.

[0058] In this embodiment, by reading the identified ear tag information, it is determined whether the livestock is to be tested based on the measurement parameters. After it is determined to be the livestock to be tested, image data is obtained. According to the backfat measurement point recognition model in the local storage, the position information and size of the target livestock in the image are identified. Combined with the actual P2 point position of the livestock's backfat, the data is comprehensively calculated and analyzed to obtain the coordinate data of the P2 point to be measured on the back of the livestock. The image of the additional coordinate point is stored in the local storage database. According to the coordinate point information, an instruction is sent to the multi-axis robotic arm motion module to drive the robotic arm carrying the backfat probe to reach the P2 point to be measured after dipping in the coupling agent. The video information is received from the backfat video data acquisition module and stored in the local storage database. It can be viewed and processed on the terminal data platform.

[0059] Among them, the unique identity information in the electronic ear tag corresponds to the age, weight, and P2 point image of the livestock's back to be measured in the local storage database. The electronic ear tag reader is set in the detection area, and the read electronic ear tag information is analyzed. If the electronic ear tag information is received for the first time, the information recorded in the electronic ear tag is filled into the livestock back fat information database and a new data item is formed.

[0060] In a possible embodiment, in step S2, the training process of the backfat measurement point recognition model includes:

[0061] Acquiring back image data of the acquisition object;

[0062] Annotating the image data of the back of the collected object;

[0063] Load the YoLov4-tiny model for model training;

[0064] Use the test set to test the trained model;

[0065] Evaluate the test results and end the training if they meet the set requirements.

[0066] In this embodiment, in step S3, the main control cabinet compiles the collected positioning coordinates and distance values ​​into corresponding instructions and sends them to the multi-axis robotic arm. After receiving the instructions, the multi-axis robotic arm first goes to the coupling agent container according to the preset trajectory to dip the coupling agent.

[0067] In this embodiment, in step S4, the multi-axis robotic arm is controlled to move at a second speed until it fits the skin of the collection object, and then the multi-axis robotic arm is continued to be controlled to move so that the skin of the collection object that is in contact with the collection object is pressed down a preset distance; it is determined whether the position of the collection object has moved. If the position of the collection object has moved, the position of the multi-axis robotic arm is adjusted according to the movement of the collection object so that the back fat B-ultrasound probe continues to remain in complete contact with the back of the collection object.

[0068] Specifically, after the coupling agent is dipped, it continues to move towards the predetermined coordinates transmitted from the main control cabinet until the distance sensor on the multi-axis robotic arm detects that the backfat B-ultrasound probe installed on the top of the multi-axis robotic arm is about to reach the point on the pig's back where backfat measurement is required. The multi-axis robotic arm starts to move slowly until it fits the pig's skin. However, in order to ensure that the backfat B-ultrasound probe is completely in contact with the pig's back, the multi-axis robotic arm will continue to press down a certain distance. If the pig moves during the measurement work, the distance camera and distance sensor will transmit the information to the main control cabinet according to the pig's movement. The main control cabinet will send corresponding instructions to move the multi-axis robotic arm to ensure that the backfat B-ultrasound probe continues to be completely in contact with the pig's back.

[0069] In this embodiment, in step S5, when the backfat B-ultrasound data are collected a set number of times, the backfat B-ultrasound probe is controlled to stop data collection; if the collection object changes the feeding position, the backfat measurement and collection of the collection object is performed again.

[0070] Specifically, when the backfat ultrasound probe reaches the backfat measurement point, it begins collecting backfat ultrasound data and transmits the collected ultrasound data to the main control cabinet in the form of a video. The backfat ultrasound probe takes approximately 20 seconds to collect data once. After the backfat ultrasound probe completes its initial data collection, as long as the electronic ear tag reader detects the pig's electronic ear tag signal, indicating that the pig is still feeding, the main control cabinet will continue to repeat the measurement work, and the backfat ultrasound probe will continue to work. The purpose of performing multiple measurements on the same pig is to cross-verify possible errors or misidentifications. If there is any doubt about the final ultrasound image collected, the retained multiple measurement image information can be verified.

[0071] In this embodiment, in step S5, the collected B-ultrasound data is stored locally, and a chart including the measurement date, ear tag number, and B-ultrasound image content is generated.

[0072] Specifically, the measurement is repeated up to 4 times during the pig's feeding period. If the same pig is still feeding after 4 measurements, the data collection work will be suspended. If 4 measurements have been reached or 4 measurements have not been reached, and a different pig is replaced to feed, the back fat measurement and collection of the newly feeding pig will be carried out.

[0073] In this embodiment, when the backfat ultrasound probe collects an ultrasound image, the image will be sent to the main control cabinet in the form of a video. The main control cabinet will save the image video in the local memory and automatically generate a chart containing the measurement date, ear tag number, B-ultrasound image and other contents.

[0074] The main control cabinet is connected to the local network via a wired or wireless connection. This allows users in the office to access data through the client software, using a computer connected to the local network. This allows users to add charts and graphs to pig growth information, such as weight and length. Backfat ultrasound images are stored locally, and the earliest images are automatically overwritten if the memory is full.

[0075] In summary, the present invention uses an electronic ear tag reader to identify the electronic ear tag on the ear of the collection object during the collection object's eating process, and obtains the number of the electronic ear tag of the collection object; after the main control cabinet receives the number of the electronic ear tag of the collection object, it starts the depth camera to collect images of the collection object, and uses the backfat measurement point recognition model to coordinate the backfat measurement point on the collected image; according to the coordinate positioning value and the distance value from the depth camera to the back of the collection object, a robot arm control instruction is generated, and the robot arm control instruction is sent to the multi-axis robot arm, and the multi-axis robot arm controls the robot arm according to the robot arm control instruction. The control instruction moves to the coupling agent container according to a preset trajectory to dip the coupling agent; after dipping the coupling agent, the multi-axis robotic arm is controlled to move to a given coordinate position at a first speed until the distance sensor on the multi-axis robotic arm detects that the backfat B-ultrasound probe installed on the top of the multi-axis robotic arm reaches the point on the back of the collection object where backfat measurement is required, and then the multi-axis robotic arm is controlled to move at a second speed until it fits the skin of the collection object; after the backfat B-ultrasound probe reaches the backfat measurement point, the backfat B-ultrasound data is repeatedly collected according to the set number of times, and the collected B-ultrasound data is transmitted to the main control cabinet in the form of video. The present invention reduces the human-animal contact link. After the pig enters the measurement station or feeding station, it is identified by an electronic ear tag and positioned by a depth camera. Then, a multi-axis robotic arm is used to move the backfat B-ultrasound probe for detection. A series of actions are performed autonomously by the machine without manual operation, which greatly reduces the risk of infection with infectious diseases such as African swine fever caused by frequent contact between humans and animals; precise positioning and efficient measurement; the trained system can accurately identify and locate, which is more secure in comparison; the stress response of the pig is reduced. In the past, backfat measurement required the pig to be driven to an unfamiliar fixed location. The driving of people and the unfamiliar environment are both possible factors that cause stress in the pig. The collection work of the present invention is carried out when the pig is feeding in a familiar environment. The pig's resistance is relatively very small and the degree of cooperation is relatively high; data storage is more convenient and complete. In the past, backfat measurement work was often done manually, and the data records were not kept in detail. The measured B-ultrasound image information could not be saved or could not be saved completely or took a longer period of time. The present invention can realize automatic saving and automatically generate reports for personnel to review.

[0076] In a possible embodiment, a sliding spray pipe is installed on the side of the backfat B-ultrasound probe. Before the backfat B-ultrasound probe performs measurement, the coupling agent in the pipe is mechanically squeezed to spray out, and the sliding decoration is used to evenly apply the coupling agent to the backfat B-ultrasound probe.

[0077] In one possible embodiment, the main function of the distance sensor is to enable the backfat B-ultrasound probe to fit better with the pig's back. This purpose can be achieved using a pressure sensor. When the pressure sensor senses that the pressure of the backfat B-ultrasound probe is a certain value, it is considered that the backfat B-ultrasound probe has formed a complete fit with the pig's back. Therefore, sensors of this type such as pressure sensors can also realize the related functions of distance sensors.

[0078] In a possible embodiment, the electronic ear tag and the electronic ear tag reader can be replaced with a graphic tag, which is pasted or otherwise fixed to the pig's body. The device is activated by recognizing the graphic tag through a camera, and the graphic tag is used to encode and record the pig's number.

[0079] In one possible embodiment, the terminal data platform allows users to view and manage the database data of the automatic backfat collection system, including electronic ear tag information, P2 point recognition image data, and backfat video data. It also allows users to enter age, weight, and backfat value data from backfat image measurements, set measurement parameters, and export data. Measurement parameters include the total number of measurements per day for the current livestock, the number of consecutive single measurements, the total number of coupling agent uses, and the number of coupling agent remaining alarms.

[0080] FIG3 is an application scenario of the method according to an embodiment of the present invention, wherein:

[0081] Client: View and manage the image data and video data of the backfat automatic acquisition system, and measure the backfat value of the backfat image.

[0082] Multi-axis robotic arm: A backfat probe and distance sensor are installed at the end, and it receives instructions to move to the point to be measured.

[0083] Distance sensor: measures the distance between the backfat probe and the point to be measured, so that the probe carried by the robotic arm can be close to the skin of the point to be measured.

[0084] Backfat probe: emits ultrasonic waves and receives ultrasonic signals, converts them into electrical signals and transmits them to the backfat main board.

[0085] Binocular depth camera: obtains color images and depth values ​​of the livestock's back.

[0086] Electronic ear tag reader: detects whether livestock have entered a specific location and obtains ear tag identity information.

[0087] Electronic ear tags: identification tags for livestock.

[0088] Control cabinet: Contains the mainboard and supporting modules of the automatic backfat collection system.

[0089] The control cabinet is equipped with a system power supply, a control mainboard, a local storage unit (SSD solid state drive), a switch, a backfat meter host, a USB video capture card, a relay module, and a USB to RS485 module.

[0090] System power supply: Provides various voltage requirements for the entire system.

[0091] Control main board: receives sensor information and processes and issues control instructions.

[0092] Local storage unit (SSD solid-state drive): stores images captured by the binocular depth camera when calculating the P2 point, images captured by the binocular depth camera when measuring backfat, and video recordings captured by the backfat meter when measuring.

[0093] Switch: controls the forwarding of data between the mainboard, multi-axis robotic arm, and the client.

[0094] Backfat meter host: receives data from the backfat probe, processes it and outputs video recording.

[0095] USB video capture card: performs interface conversion on the video signal of the backfat meter host.

[0096] Relay module: controls the start-up power supply of the backfat meter and shuts down the meter after the measurement is completed to avoid polarization damage to the backfat probe caused by prolonged electrification.

[0097] USB to RS485 module: converts the signal interface of the electronic ear tag reader.

[0098] Example 2

[0099] Referring to FIG4 , embodiment 2 of the present invention provides a backfat collection system for animals, comprising:

[0100] The ear tag number acquisition module 01 is used to identify the electronic ear tag on the ear of the collection subject through the electronic ear tag reader while the collection subject is eating, and obtain the number of the electronic ear tag of the collection subject;

[0101] Image acquisition module 02, configured to activate a depth camera to acquire an image of the object after the main control cabinet receives the number of the electronic ear tag of the object;

[0102] A coordinate positioning module 03 is used to coordinately locate the backfat measurement point on the captured image using a backfat measurement point recognition model;

[0103] A robot arm control instruction generating module 04 is configured to generate a robot arm control instruction based on the coordinate positioning value and the distance value from the depth camera to the back of the captured object;

[0104] The coupling agent dipping control module 05 is used to send the robot arm control instruction to the multi-axis robot arm, and the multi-axis robot arm moves to the coupling agent container along a preset trajectory to dip the coupling agent according to the robot arm control instruction;

[0105] The robotic arm motion control module 06 is configured to, after the coupling agent is applied, control the multi-axis robotic arm to move to a given coordinate position at a first speed until a distance sensor on the multi-axis robotic arm detects that a backfat B-ultrasound probe mounted on the top of the multi-axis robotic arm has reached a point on the back of the subject where backfat measurement is required, and then control the multi-axis robotic arm to move at a second speed until it contacts the subject's skin.

[0106] The backfat B-ultrasound data acquisition module 07 is used to repeatedly acquire backfat B-ultrasound data according to a set number of times after the backfat B-ultrasound probe reaches the backfat measurement point, and transmit the acquired B-ultrasound data to the main control cabinet in the form of video.

[0107] In this embodiment, a model training module 08 is further included for training the backfat measurement point recognition model; the model training module 8 includes:

[0108] An image loading submodule 81 is used to obtain image data of the back of the subject;

[0109] An image data annotation submodule 82 is used to annotate the image data of the back of the captured object;

[0110] The model loading submodule 83 is used to load the YoLov4-tiny model for model training;

[0111] A model testing submodule 84 is used to test the trained model using a test set;

[0112] The test evaluation submodule 85 is used to evaluate the test results and end the training if the set requirements are met.

[0113] In this embodiment, the robotic arm motion control module 06 is also used to control the multi-axis robotic arm to move at a second speed until it fits the skin of the collection object, and then continue to control the movement of the multi-axis robotic arm to press the skin of the collection object down a preset distance.

[0114] This embodiment also includes a robotic arm position adjustment module 09, which is used to determine whether the position of the collection object has moved. If the position of the collection object has moved, the position of the multi-axis robotic arm is adjusted according to the movement of the collection object so that the back fat B-ultrasound probe continues to be fully in contact with the back of the collection object.

[0115] In this embodiment, in the backfat B-ultrasound data acquisition module 07, when the backfat B-ultrasound data is collected a set number of times, the backfat B-ultrasound probe is controlled to stop data collection; if the collection object changes the feeding position, the backfat measurement and collection of the collection object is performed again;

[0116] It also includes a data storage module 10 for locally storing the collected B-ultrasound data and generating a chart containing the measurement date, ear tag number, and B-ultrasound image content.

[0117] It should be noted that the information interaction, execution process and other contents between the modules of the above-mentioned system are based on the same concept as the method embodiment in Example 1 of this application, and the technical effects they bring are the same as those of the method embodiment of this application. For specific contents, please refer to the description in the method embodiment shown above in this application, and no further details will be given here.

[0118] Example 3

[0119] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium, in which a program code of a veterinary back fat collection method is stored. The program code includes instructions for executing the veterinary back fat collection method of embodiment 1 or any possible implementation thereof.

[0120] Computer-readable storage media can be any available medium that can be accessed by a computer, or a data storage device such as a server or data center that includes one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

[0121] Example 4

[0122] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor;

[0123] The processor and the memory communicate with each other via a bus; the memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the veterinary back fat collection method of Example 1 or any possible implementation thereof.

[0124] Specifically, the processor can be implemented by hardware or by software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc.; when implemented by software, the processor can be a general-purpose processor, which is implemented by reading software code stored in a memory. The memory can be integrated into the processor or located outside the processor and exist independently.

[0125] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, computer, server or data center to another website, computer, server or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode.

[0126] Obviously, those skilled in the art will appreciate that the various modules or steps of the present invention described above can be implemented using a general-purpose computing device, centralized on a single computing device, or distributed across a network of multiple computing devices. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. In some cases, the steps shown or described can be performed in a different order than that shown, or can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0127] Although the present invention has been described in detail above using general descriptions and specific embodiments, it will be apparent to those skilled in the art that modifications and improvements may be made thereto. Therefore, such modifications and improvements, without departing from the spirit of the present invention, are intended to be within the scope of protection claimed herein.

Claims

1. A method for collecting back fat for animals, characterized in that: include: S1. When the subject is eating, the electronic ear tag on the ear of the subject is identified by an electronic ear tag reader / writer to obtain the number of the electronic ear tag of the subject; S2. When the main control cabinet receives the number of the electronic ear tag of the acquisition object, it starts the depth camera to acquire an image of the acquisition object, and uses the backfat measurement point recognition model to locate the coordinates of the backfat measurement point on the acquired image; S3, generating a robot arm control instruction according to the coordinate positioning value and the distance value from the depth camera to the back of the acquisition object, sending the robot arm control instruction to the multi-axis robot arm, and the multi-axis robot arm moves to the coupling agent container according to the preset trajectory according to the robot arm control instruction to dip the coupling agent; S4, after the coupling agent is applied, the multi-axis robotic arm is controlled to move to a given coordinate position at a first speed until a distance sensor on the multi-axis robotic arm detects that a back fat B-ultrasound probe installed on the top of the multi-axis robotic arm reaches a point on the back of the collection object where back fat measurement is required, and the multi-axis robotic arm is controlled to move at a second speed until it fits the skin of the collection object; S5. When the backfat B-ultrasound probe reaches the backfat measurement point, the backfat B-ultrasound data is repeatedly collected according to a set number of times, and the collected B-ultrasound data is transmitted to the main control cabinet in the form of video.

2. A method for collecting backfat for animals according to claim 1, characterized in that: In step S2, the training process of the backfat measurement point recognition model includes: Acquiring back image data of the acquisition object; Annotating the image data of the back of the acquisition object; Load the YoLov4-tiny model for model training; Use the test set to test the trained model; Evaluate the test results and end the training if they meet the set requirements.

3. The method for collecting back fat for animals according to claim 1, characterized in that: In step S4, the multi-axis robot arm is controlled to move at a second speed until it fits the skin of the collection object, and then the multi-axis robot arm is continued to be controlled to move so that the fitted skin of the collection object is pressed down by a preset distance.

4. A method for collecting back fat for animals according to claim 2, characterized in that: Step S4 also includes determining whether the position of the collection object moves. If the position of the collection object moves, adjusting the position of the multi-axis robotic arm according to the movement of the collection object so that the back fat B-ultrasound probe continues to be fully fitted with the back of the collection object.

5. The method for collecting backfat for animals according to claim 1, characterized in that: In step S5, when the backfat B-ultrasound data are collected for a set number of times, the backfat B-ultrasound probe is controlled to stop data collection; if the collection object changes the feeding position, the backfat measurement and collection of the collection object is performed again; In step S5, the collected B-ultrasound data is stored locally, and a chart including the measurement date, ear tag number, and B-ultrasound image content is generated.

6. A backfat collection system for animals, characterized in that: include: An ear tag number acquisition module is used to identify the electronic ear tag on the ear of the collection object through an electronic ear tag reader / writer during the collection object's eating process, and obtain the number of the electronic ear tag of the collection object; The image acquisition module is used to start the depth camera to acquire images of the acquisition object after the main control cabinet receives the number of the electronic ear tag of the acquisition object; A coordinate positioning module, used for positioning the backfat measurement points on the collected image using a backfat measurement point recognition model; A robot arm control instruction generation module, used to generate a robot arm control instruction according to the coordinate positioning value and the distance value from the depth camera to the back of the acquisition object; A coupling agent dipping control module, used for sending the robot arm control instruction to the multi-axis robot arm, and the multi-axis robot arm moves to the coupling agent container along a preset trajectory to dip the coupling agent according to the robot arm control instruction; A mechanical arm motion control module, used for controlling the multi-axis mechanical arm to move to a given coordinate position at a first speed after the coupling agent is applied, until a distance sensor on the multi-axis mechanical arm detects that a back fat B-ultrasound probe installed on the top of the multi-axis mechanical arm reaches a point on the back of the collection object where back fat measurement is required, and then controlling the multi-axis mechanical arm to move at a second speed until it fits against the skin of the collection object; The backfat B-ultrasound data acquisition module is used for, when the backfat B-ultrasound probe reaches the backfat measurement point, The backfat B-ultrasound data is repeatedly collected according to the set number of times, and the collected B-ultrasound data is transmitted to the main control cabinet in the form of video.

7. The animal backfat collection system according to claim 6, characterized in that: It also includes a model training module for training the backfat measurement point recognition model; the model training module includes: An image loading submodule, used for acquiring back image data of the acquisition object; An image data annotation submodule, used to annotate the image data of the back of the acquisition object; Model loading submodule, used to load YoLov4-tiny model for model training; The model testing submodule is used to test the trained model using the test set; The test evaluation submodule is used to evaluate the test results and end the training if the set requirements are met.

8. The animal backfat collection system according to claim 6, characterized in that: The robot arm motion control module is also used to control the multi-axis robot arm to move at a second speed until it fits against the skin of the collection object, and then continue to control the multi-axis robot arm to move so that the fitted skin of the collection object is pressed down a preset distance.

9. The animal backfat collection system according to claim 7, characterized in that: It also includes a robotic arm position adjustment module, which is used to determine whether the position of the collection object moves. If the position of the collection object moves, the multi-axis robotic arm is adjusted in position according to the movement of the collection object so that the back fat B-ultrasound probe continues to be fully fitted with the back of the collection object.

10. The animal backfat collection system according to claim 6, characterized in that: In the backfat B-ultrasound data acquisition module, when the backfat B-ultrasound data is collected for a set number of times, the backfat B-ultrasound probe is controlled to stop data collection; if the collection object changes the feeding position, the backfat measurement and collection of the collection object is performed again; It also includes a data storage module for locally storing the collected B-ultrasound data and generating a chart containing the measurement date, ear tag number, and B-ultrasound image content.

Images