阅读说明：本技术 一种肉牛健康自动监测系统 (Beef cattle health automatic monitoring system ) 是由 张东彦 李威风 张淦 黄小平 林志 王鑫 于 2021-08-19 设计创作，主要内容包括：本发明特别涉及一种肉牛健康自动监测系统,包括采集通道和处理模块,所述的采集通道仅供单头牛进入,且设置有摄像头用于拍摄牛脸照片或视频,设置有压力和温度传感器用于采集牛的体重和体温；处理模块接收牛脸照片或视频后对单头牛进行识别并将传感器获取的数据与识别的牛的身份信息相关联,进而对牛的体重和体温数据进行分析与处理得到牛的健康状况并输出。其中,采集通道每次只能一头牛进入采集其相关数据,确保牛场中所有牛的数据都是独立的；此外,本发明通过拍摄牛场中每头牛的脸来对其身份进行识别,无需在耳朵上安装耳标,避免了背景技术中述及的诸多不足。(The invention particularly relates to an automatic beef cattle health monitoring system which comprises an acquisition channel and a processing module, wherein the acquisition channel is only used for single cattle to enter, is provided with a camera for shooting cattle face pictures or videos, and is provided with pressure and temperature sensors for acquiring the weight and the body temperature of the cattle; the processing module receives the cattle face pictures or videos, identifies the single cattle, associates the data acquired by the sensor with the identified cattle identity information, and further analyzes and processes the weight and body temperature data of the cattle to obtain the health condition of the cattle and outputs the health condition. Wherein, the acquisition channel can only allow one cow to enter and acquire related data each time, so that the data of all the cows in the cow farm are independent; in addition, the invention identifies the identity of each cow by shooting the face of each cow in a cattle farm without installing ear tags on ears, thereby avoiding the defects in the background technology.)

1. The utility model provides a beef cattle health automatic monitoring system which characterized in that: the cattle body weight acquisition system comprises an acquisition channel and a processing module, wherein the acquisition channel is only used for single cattle to enter, is provided with a camera (500) for shooting cattle face pictures or videos, and is provided with pressure and temperature sensors for acquiring the weight and the body temperature of the cattle; the processing module receives the cattle face pictures or videos, identifies the single cattle, associates the data acquired by the sensor with the identified cattle identity information, and further analyzes and processes the weight and body temperature data of the cattle to obtain the health condition of the cattle and outputs the health condition.

2. The automatic beef cattle health monitoring system of claim 1, wherein: the collecting channel is arranged in the cattle pen and comprises rails (100), weighbridges (200) and food troughs (300), the food troughs (300) are arranged on one side edge of the cattle pen, two rails (100) are arranged in parallel and are perpendicular to the edge of the cattle pen placing food trough (300); the space between the two fences (100) is matched with the width of the trough (300), and the space is larger than the body width of one cow and smaller than the body width of two cows; the ground between the two fences (100) is provided with a wagon balance (200) for collecting the weight of the cattle.

3. The automatic beef cattle health monitoring system of claim 2, wherein: crib (300) wholly be squarely, its upper end face has seted up the notch downwards and is used for holding water or foodstuff, one side that fence (100) were kept away from in crib (300) is provided with camera (500), infrared temperature probe (600) and drive unit (400), camera (500) are used for shooting ox face photo, infrared temperature probe (600) are used for detecting the body temperature of ox, drive unit (400) are used for driving camera (500) and infrared temperature probe (600) and move to the dead ahead of ox face.

4. An automatic beef cattle health monitoring system as claimed in claim 3, wherein: the processing module comprises a microprocessor installed in the cattle pen and a computer installed outside the cattle pen, and the microprocessor acquires data according to the following steps:

A. the microprocessor collects output data of the wagon balance (200) at intervals of t1, if the data of the wagon balance (200) is more than 0, the next step is executed, otherwise, the step is repeated;

B. the microprocessor controls the camera (500) to start and take a picture, and the microprocessor processes the picture to obtain the position of the cow face;

C. the microprocessor controls the driving unit (400) to start according to the current position of the camera (500) and the cow face position;

D. the driving unit (400) drives the camera (500) and the infrared temperature measuring probe (600) to move to the front of the cow face;

E. the microprocessor collects output data of the wagon balance (200) at intervals of t2, if the data of the wagon balance (200) is 0, step F is executed, otherwise step E is repeated; in the process of repeatedly executing the step E, the microprocessor controls the camera (500) to take a picture at intervals of t3, controls the infrared temperature measuring probe (600) to acquire the body temperature of the cow and stores all data;

F. and C, packaging and sending all the stored data to the computer by the microprocessor and returning to the step A.

5. The automatic beef cattle health monitoring system of claim 4, wherein: in the step E, "in the process of repeatedly executing the step E, the microprocessor controls the camera (500) to take a picture and store the picture at intervals of time t 3" to replace: the microprocessor controls the camera (500) to shoot the video and store the video in the process of repeatedly executing the step E; and step E, acquiring the infrared image of the cow by the infrared temperature measuring probe (600) and storing the infrared image.

6. The automatic beef cattle health monitoring system of claim 5, wherein: the computer processes the received data according to the following steps:

s100, recognizing the cattle face picture or video in one data packet to obtain the identity information of the cattle;

s200, calculating the weight of the cattle and the feed intake according to the collected discrete cattle weight data;

s300, processing the collected infrared images of the plurality of cattle to obtain the body temperature of the cattle;

s400, drawing a weight curve, a feed intake curve and a body temperature curve of the cattle according to the data acquisition time, the identity information of the cattle, the weight of the cattle, the feed intake of the cattle and the body temperature of the cattle;

s500, comparing the weight curve and the body temperature curve of the cattle with a standard curve, and sending out a health early warning when the difference value exceeds a set threshold value; and calculating the meat-feed ratio according to the weight and the feed intake of the cattle, and giving out a health early warning when the meat-feed ratio data is abnormal.

7. The automatic beef cattle health monitoring system of claim 6, wherein: the step S200 includes the following steps:

s201, establishing the following equation:

wherein a and b are weight constants to be solved, t_a、t_bIs the time constant to be solved;

s202, recording discrete cow weight data as { (y1, t1), (y2, t2), …, (yn, tn) };

s203, calculating the nested loop according to the following formula_ijWherein, the value of i of the outer circulation is {1, 2, …, n-1}, the value of j of the inner circulation is { i +1, i +2, …, n },

s204, solving S_ijAnd taking a and b corresponding to the minimum value, wherein a is the weight of the cattle, and b is the feed intake of the cattle.

8. The automatic beef cattle health monitoring system of claim 7, wherein: the step 300 includes the following steps:

s301, setting the reference lowest temperature of the forehead of the cow as T1, the reference highest temperature as T2, the lower limit of the reference area as A1 and the upper limit as A2;

s302, performing threshold segmentation and connected domain analysis on each frame of infrared image, judging that the forehead region is a region when the temperature of a connected region is greater than T1 and less than T2 and the area of the connected region is greater than A1 and less than A2, and taking the maximum temperature value in the region as the cattle body temperature T of the infrared image of the frame_nWherein n is the number of frames;

s303, analyzing all the infrared images to obtain a plurality of ox body temperature data { T }₁，T₂，…，T_NGet the maximum value T_MAXAs the body temperature of a single cow.

9. The automatic beef cattle health monitoring system of claim 8, wherein: in the step S100, a MTCNN face recognition technology is used to build a deep learning network model, which specifically includes: s101, performing coarse positioning and accurate positioning on the bovine face target by respectively using P net and R net; s102, positioning the center of gravity of the five sense organs of the cattle by using O net; s103, recognizing the face of the cow according to the characteristics of the five sense organs of the cow, and confirming the identity information of the cow;

in the step S500, the standard weight curve and the standard body temperature curve of the cattle are obtained by averaging the weight data and the body temperature data of healthy cattle, and the meat-to-feed ratio is calculated according to the following formula: and P is delta W/delta M, wherein delta W is the change value of the weight of the cattle systematically recorded in the designated time, and delta W is the sum of the feed intake of the cattle systematically recorded in the designated time.

10. The automatic beef cattle health monitoring system of claim 9, wherein: the fence (100) is in a square grid shape; the driving unit (400) comprises a first motor (401), a screw rod (402), a sliding rod (403), a moving support rod (404), a second motor (405), a belt (406) and a mounting seat (407), the screw rod (402) and the sliding rod (403) are arranged in the vertical direction, the shaft cores of the screw rod (402) and the sliding rod (403) are respectively located in the planes of two fences (100), the upper ends of the screw rod (402) and the sliding rod (403) are fixedly mounted on a protruding part extending out of the upper end of the fence (100), the lower ends of the screw rod (402) and the sliding rod (403) are fixedly mounted on the crib (300), the moving support rod (404) is horizontally mounted on the screw rod (402) and the sliding rod (403), the moving support rod (404) moves up and down along the length direction of the screw rod (402) when the screw rod (402) is driven to rotate by the first motor (401), the second motor (405) is mounted at one end of the moving support rod (404), and the mounting seat (407) is driven by the belt (406) to move horizontally along the length direction of the moving support rod (404), the camera (500) and the infrared temperature measuring probe (600) are both arranged on the mounting seat (407).

Technical Field

The invention relates to the technical field of beef cattle breeding, in particular to an automatic beef cattle health monitoring system.

Background

Beef has high protein content, low fat content and comprehensive nutrition, and occupies larger and larger proportion in meat consumption, and the consumption demand of milk and milk products is continuously increased. With the growth of cattle raising industry, the raising mode is large-scale and has higher and higher intensification degree, intensive breeding is implemented in the current cattle farm, cattle are limited in a cattle shed in the growth process and lack of activity, so that the taste of produced beef is reduced, the quality of the milk is reduced, air flow in the cattle shed is unsmooth, the sanitary condition is poor, diseases are easily caused, and the healthy growth of the cattle is not facilitated. The use of additives such as antibiotics, hormones, clenbuterol and the like in the beef feed influences the beef quality; the poor culture environment causes diseases such as respiratory tract, digestive tract, mastitis and the like, and the use of antibiotics influences the food safety of milk and products thereof.

In order to solve the problems, more and more farms start information-based and intelligent breeding, the physical conditions of the cattle are monitored by installing ear tags and a plurality of sensors on the cattle, the ear tags are used for distinguishing the cattle, the existing ear tags are mostly fixed on the ears of the cattle and are internally provided with RFID tags, and when the cattle are close to a collecting device, the sensors read the information of the body conditions of the cattle, which is the information of the ear tags, and process the information. This solution has a number of disadvantages: firstly, the ear tag is fixed on the ear of a cow, often falls off, is not reliable enough in fixation, and needs to be confirmed by a farmer often, otherwise, the information of the cow is lost in the period of time; secondly, the ear tag can cause certain harm to the ears of the cattle, and the ears of the cattle fixed with the ear tag are basically torn; thirdly, when carrying out ox information reading, sometimes there is the bull near, and the ear tag of other cattle can influence the accuracy that data read, can appear associating the information that the sensor was gathered to other cattle promptly, and then cause the data error. Above, all can influence the accuracy and the reliability of ox data acquisition.

Disclosure of Invention

The invention aims to provide an automatic beef cattle health monitoring system which can reliably and accurately acquire cattle information.

In order to realize the purpose, the invention adopts the technical scheme that: an automatic beef cattle health monitoring system comprises an acquisition channel and a processing module, wherein the acquisition channel is only used for single cattle to enter, is provided with a camera for shooting cattle face pictures or videos, and is provided with pressure and temperature sensors for acquiring the weight and the body temperature of the cattle; the processing module receives the cattle face pictures or videos, identifies the single cattle, associates the data acquired by the sensor with the identified cattle identity information, and further analyzes and processes the weight and body temperature data of the cattle to obtain the health condition of the cattle and outputs the health condition.

Compared with the prior art, the invention has the following technical effects: only one cow can enter the acquisition channel to acquire related data each time, so that the data of all cows in the cow farm are independent; meanwhile, the identity of each cow in the cow farm is recognized by shooting the face of each cow, ear tags do not need to be installed on ears, and many defects in the background art are avoided.

Drawings

FIG. 1 is a schematic structural view of the present invention;

FIG. 2 is another angular configuration of FIG. 1;

fig. 3 is a schematic diagram of a cow face identification process.

Detailed Description

The present invention will be described in further detail with reference to fig. 1 to 3.

Referring to fig. 1 and 2, a beef cattle health automatic monitoring system, including gathering passageway and processing module, the collection passageway supply the single entering of ox, be provided with camera 500 in the collection passageway and be used for taking the face photo of ox or video, still be provided with the sensor in the collection passageway and be used for gathering the weight and the body temperature of ox, processing module discerns the ox after receiving face photo of ox or video and associates the data that the sensor was gathered with the identity information of the ox of discernment, processing module still handles the health status and the output that obtain the ox to the weight and the body temperature data of ox. Only one cow can enter the acquisition channel to acquire related data each time, so that the data of all cows in the cow farm are independent; meanwhile, the identity of each cow in the cow farm is recognized by shooting the face of each cow, ear tags do not need to be installed on ears, and many defects in the background art are avoided.

The structure of the collecting channel is various, and the collecting channel is preferably arranged in the cattle pen, the collecting channel comprises a fence 100, a weighbridge 200 and a food groove 300, the food groove 300 is arranged at the edge of one side of the cattle pen, two fences 100 are arranged in parallel and are perpendicular to the edge of the cattle pen placing food groove 300; the space between the two fences 100 is matched with the width of the trough 300, and the space is larger than the body width of one cow and smaller than the body width of two cows; a weighbridge 200 is provided on the ground between the two pens 100 for weighing the cattle. By adopting the structure, the mounting and the dismounting are convenient, and the weight information of the cattle can be conveniently collected by the arrangement of the wagon balance 200.

Further, the trough 300 wholly be squarely, its upper end face has seted up the notch downwards and is used for holding water or foodstuff, one side that the fence 100 was kept away from to the trough 300 is provided with camera 500, infrared temperature probe 600 and drive unit 400, camera 500 is used for shooting the ox face photo, infrared temperature probe 600 is used for detecting the body temperature of ox, drive unit 400 is used for driving camera 500 and infrared temperature probe 600 and moves to the dead ahead of ox face. Camera 500's setting, the photo and the video of collection ox face that can be convenient to make things convenient for subsequent ox face to discern, and then can be associated the data of gathering with the identity of ox. The temperature of the cattle can be conveniently detected by arranging the infrared temperature measuring probe 600. The shooting process and the temperature measuring process are quiet, so that the cattle are prevented from being threatened to influence the feeding.

Further, the processing module include install the microprocessor in the cowshed and install the computer outside the cowshed, after setting up like this, distinguish data acquisition and processing, at big plant like this, can all set up a microprocessor on every collection passageway, the data that all microprocessors gathered all transmit to the computer through wired or wireless mode, the computer is unified to all data and is handled, because microprocessor only carries out the collection and the packing storage of data, so the cost can be done very lowly.

Further, the microprocessor performs data acquisition according to the following steps: A. the microprocessor collects output data of the wagon balance 200 at intervals of T1, if the data of the wagon balance 200 is greater than 0, a cow enters the wagon balance, and a value greater than 0 can be set during actual operation, so that misjudgment caused by floating of the wagon balance 200 data is avoided, the next step is executed, and otherwise, the step is repeated; B. the microprocessor controls the camera 500 to start and take a picture, and the microprocessor processes the picture to obtain the position of the cow face; C. the microprocessor controls the driving unit 400 to start according to the current position of the camera 500 and the cow face position; D. the driving unit 400 drives the camera 500 and the infrared temperature measuring probe 600 to move to the position right in front of the cow face, and the step B, C, D is executed in the process that the cow enters the acquisition channel, namely, the cow entering is detected, the approximate position of the face of the cow is estimated immediately, and then the positions of the camera 500 and the infrared temperature measuring probe 600 are adjusted to the position right in front of the cow face, wherein the position processing of the cow face can be simple processing and is not required to be particularly accurate; E. the microprocessor collects output data of the wagon balance 200 at intervals of T2, if the data of the wagon balance 200 is 0, the cattle leaves, the step F is executed, otherwise, the cattle is still in the collection channel, and the step E is repeated; in the process of repeatedly executing the step E, the microprocessor controls the camera 500 to take a picture at intervals of T3, controls the infrared temperature measuring probe 600 to acquire the body temperature of the cow and stores all data; F. and C, packaging and sending all the stored data to the computer by the microprocessor and returning to the step A.

Further, in the step E, "in the process of repeatedly executing the step E, the microprocessor controls the camera 500 to take a picture and store the picture at intervals of time T3" instead of: and E, repeatedly executing the step E, wherein the microprocessor controls the camera 500 to shoot the video and store the video. That is, when the ox passes through the acquisition channel, the shooting of the ox face photo is removed, the video can be directly shot for processing, the photo is intermittently acquired, the video can be understood as the photo acquired at high frequency, the data volume is more, the identity information of the ox is more accurately analyzed, and the consumed flow during the transmission is more. In the step E, the infrared temperature probe 600 collects and stores infrared images of the cow, and the infrared temperature probe does not directly collect the body temperature of the cow, because the difference of temperature measurement points causes a large temperature error, and the more accurate body temperature of the cow can be obtained by collecting the infrared images of the cow and then processing the infrared images in the computer.

The computer may be an industrial personal computer or other processing units, and generally, software with corresponding functions is installed on the computer to realize functions of data storage, analysis and the like. The computer processes the received data according to the following steps: s100, recognizing the cattle face picture or video in one data packet to obtain the identity information of the cattle; s200, calculating the weight of the cattle and the feed intake according to the collected discrete cattle weight data; s300, processing the collected infrared images of the plurality of cattle to obtain the body temperature of the cattle; s400, drawing a weight curve, a feed intake curve and a body temperature curve of the cattle according to the data acquisition time, the identity information of the cattle, the weight of the cattle, the feed intake of the cattle and the body temperature of the cattle; s500, comparing the weight curve and the body temperature curve of the cattle with a standard curve, and sending out a health early warning when the difference value exceeds a set threshold value; and calculating the meat-feed ratio according to the weight and the feed intake of the cattle, and giving out a health early warning when the meat-feed ratio data is abnormal. Only the weight, the body temperature and the meat-to-material ratio of the cattle are monitored and early-warned, and an operator can set more sensors, collect more other data of the cattle and analyze more data according to the requirement so as to realize more comprehensive early warning.

In order to accurately calculate the weight and food intake of the cow, in the present invention, preferably, the step S200 includes the following steps: s201, establishing the following equation:

wherein a and b are weight constants to be solved, t_a、t_bThe equation is established for the time constant to be solved because the weight of the cattle is theoretically constant just before the cattle enters and leaves the cattle, and the weight of the cattle is theoretically uniformly increased during the feeding process; s202, recording discrete weight data of the cattle as { y1, t1, y2, t2, …, yn, tn }; now only the best equation needs to be fitted according to these data, and the following steps can be specifically adopted: s203, calculating the nested loop according to the following formula_ijWherein, the value of i of the outer circulation is {1, 2, …, n-1}, the value of j of the inner circulation is { i +1, i +2, …, n },

s204, solving S_ijAnd taking a and b corresponding to the minimum value, wherein a is the weight of the cattle, and b is the feed intake of the cattle. By adopting nested loop calculation, the computer can quickly solve the optimal values of a and b, the calculation process is simple, and the actual calculation is very quick.

Further, in order to measure the body temperature of the cow more accurately, the step 300 includes the following steps: s301, setting the reference lowest temperature of the forehead of the cow as T1, the highest temperature as T2, the lower limit of the reference area as A1 and the upper limit as A2, wherein the thresholds are preset according to experience; s302, performing threshold segmentation and connected domain analysis on each frame of infrared image, judging that the forehead region is a region when the temperature of a connected region is greater than T1 and less than T2 and the area of the connected region is greater than A1 and less than A2, and taking the maximum temperature value in the region as the cattle body temperature T of the infrared image of the frame_nWherein n is the number of frames; can comparatively accurate elect the ox forehead region like this, the ox forehead regionThe maximum temperature value is closer to the real body temperature; s303, analyzing all the infrared images to obtain a plurality of ox body temperature data { T }₁，T₂，…，T_NGet the maximum value T_MAXAs the body temperature of the cattle. Therefore, the body temperature data of the cattle can be acquired more accurately.

Referring to fig. 3, further, in step S100, building a deep learning network model by using the MTCNN face recognition technology specifically includes: s101, performing coarse positioning and accurate positioning on the bovine face target by respectively using P net and R net; s102, positioning the center of gravity of the five sense organs of the cattle by using O net; s103, recognizing the face of the cow according to the characteristics of the five sense organs of the cow, and confirming the identity information of the cow. The MTCNN is mainly used for a model for face recognition, and the face of the cow can form the characteristics of the five sense organs by positioning the five sense organs, so that the algorithm is directly adopted for face recognition.

Further, in step S500, the standard weight curve and the standard body temperature curve of the cattle are obtained by averaging the weight data and the body temperature data of healthy cattle, which may be collected in advance and then stored in the computer. The meat-material ratio is calculated according to the following formula: and P is delta W/delta M, wherein delta W is the change value of the weight of the cattle recorded by the system in the specified time, and delta W is the sum of the feed intake of the cattle recorded by the system in the specified time.

Further, the fence 100 is in a square grid shape, so that the weight of the fence 100 can be reduced, and the installation and the disassembly are convenient. There are various structures of the driving unit 400, and it is preferable here that the driving unit 400 includes a first motor 401, a screw 402, a slide bar 403, a moving bar 404, a second motor 405, a belt 406, and a mounting seat 407, the lead screw 402 and the slide bar 403 are arranged in the vertical direction, the axial centers of the lead screw 402 and the slide bar 403 are respectively located in the surfaces of the two fences 100, the upper ends of the lead screw 402 and the slide bar 403 are fixedly installed on a protruding portion extending out of the upper end of the fence 100, the lower ends of the lead screw 402 and the slide bar 403 are fixedly installed on the crib 300, the movable support rod 404 is horizontally installed on the lead screw 402 and the slide bar 403, the movable support rod 404 moves up and down along the length direction of the lead screw 402 when the lead screw 402 is driven to rotate by the first motor 401, the second motor 405 is installed at one end of the movable support rod 404, the mounting seat 407 is driven by the second motor 405 through the belt 406 to translate along the length direction of the movable support rod 404, and the camera 500 and the infrared temperature measuring probe 600 are both installed on the mounting seat 407. The up-and-down movement of the mounting seat 407 is realized by the thread fit between the screw rod 402 and the moving support rod 404, and the left-and-right movement is realized by the belt 406, so that the device is a simple two-dimensional moving platform. Through above device, can be convenient remove camera 500 and infrared temperature probe 600.