CN114821400A - Animal body condition evaluation system and method - Google Patents
Animal body condition evaluation system and method Download PDFInfo
- Publication number
- CN114821400A CN114821400A CN202210361831.6A CN202210361831A CN114821400A CN 114821400 A CN114821400 A CN 114821400A CN 202210361831 A CN202210361831 A CN 202210361831A CN 114821400 A CN114821400 A CN 114821400A
- Authority
- CN
- China
- Prior art keywords
- information
- point
- animal
- plane
- pixel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
An animal condition scoring system comprises a depth image acquisition module and a body condition scoring module; the depth image acquisition module is used for acquiring a back depth image reflecting the back contour of the animal from the upper part of the animal to be detected and providing the back depth image for the body condition scoring module; the body condition scoring module is used for extracting image information of key parts reflecting the fat and thin of the animal and image information of a reference part from the acquired back depth image; calculating a reference fat-lean volume value and integrating an actual fat-lean volume value according to the extracted image information of the key part and the image information of the reference part; and taking the reference fat-lean volume value and the actual fat-lean volume value as a ratio, and multiplying the ratio by a coefficient to obtain the body condition score value BCS of the animal. The invention also provides a method for scoring the animal body condition.
Description
The technical field is as follows:
the invention relates to the technical field of animal health monitoring, in particular to an animal condition evaluation system and method.
Background art:
the animal Body Condition score (Body Condition Scoring), also known as fat Condition assessment, substantially reflects the amount of fat deposited in the animal and thus whether the animal is fed for a period of time to meet the requirements, in order to adjust the feeding method. For example, dairy cattle, feeding practices and scientific experiments have demonstrated that cattle body conditions can affect potential or acute post-partum complications, milk production during lactation, and reproductive efficiency. In the process of feeding the dairy cows, proper body condition grading is carried out at different stages, so that a feeder can be helped to better know the nutritional condition and the body condition of the dairy cows and the problems in the feeding management process, so that relevant measures can be taken in time to solve the problems, and the feeding level of the dairy cows can be better improved.
Body condition scoring is a special method for evaluating the body fat deposition amount of dairy cows, is an important index for evaluating the production capacity of a herd, testing and evaluating the feeding level of a dairy farm, and is an indispensable part in the breeding management of the dairy farm. Cow body condition refers to the amount of fat or reserve level of energy that a cow has. The purpose of body condition scoring is to keep the cow in a good weight state, and to clarify the problems of cows in different growth and production stages and the influence factors which may be brought to the cow by the fluctuation of body conditions.
At present, most of dairy cow body condition scores are given by observing and analyzing related parts of the dairy cow by a scoring person who is trained professionally and has related scoring experience, the score is given by 5, the low score represents emaciation, the high score represents obesity, and 0.25 is taken as a scoring unit. The traditional method is too time-consuming and is influenced by subjective factors of evaluators more, and the error is larger.
The invention content is as follows:
accordingly, there is a need for an animal condition scoring system that avoids scoring errors caused by human subjective factors.
There is also a need to provide a method for scoring animal body conditions that avoids scoring errors caused by human subjective factors.
An animal condition scoring system comprises a depth image acquisition module and a body condition scoring module; the depth image acquisition module is used for acquiring a back depth image reflecting the back contour of the animal from the upper part of the animal to be detected and providing the back depth image for the body condition scoring module;
the body condition scoring module is used for extracting image information of key parts reflecting animal fat and thinness and image information of reference parts from the acquired back depth image; the image information comprises pixel information reflecting the key part and the reference part, and a three-dimensional coordinate point corresponding to each pixel information; acquiring a predetermined amount of plane feature point information for constructing a plane according to the extracted image information of the key part and a first extraction rule, wherein the plane feature point information comprises pixel information corresponding to the plane feature point and a three-dimensional coordinate point corresponding to the pixel information; acquiring height reference characteristic point information according to the extracted image information of the reference part and a second extraction rule, wherein the height reference characteristic point information comprises pixel information corresponding to the height reference characteristic point and a three-dimensional coordinate point corresponding to the pixel information; constructing a reference cube reflecting the back slimming of the animal according to the height reference feature point information, the plane feature point information and a preset height value, wherein the upper surface and the lower surface of the reference cube are the same in shape and size, the upper surface and the lower surface are parallel and opposite, and the height reference feature point is positioned in the upper surface of the reference cube; calculating the volume of the reference cube according to the plane feature points and the preset height value, and taking the calculated volume value as a reference fat-lean volume value V whole (ii) a Calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube according to the pixel information in the target image area surrounded by the plane feature points, the three-dimensional coordinate point corresponding to the pixel information and the three-dimensional coordinate point of the lower surface of the reference cube; calculating the volume value corresponding to each pixel in the target image region according to the relative height value, and integrating the volume of each pixel in the target image region to obtain the actual fat-lean volume value V reflecting animal fat and lean reality (ii) a Substituting the reference fat-lean volume value and the actual fat-lean volume value into a formula I for calculation so as to obtain a body condition score value BCS of the animal;
wherein δ in the formula one is a correction coefficient.
A method of scoring a body condition of an animal comprising the steps of:
acquiring a back depth image reflecting the back contour of an animal from the upper part of the animal to be detected by using a depth camera;
extracting image information of key parts reflecting animal fatness and thinness and image information of reference parts from the acquired back depth image; the image information comprises pixel information reflecting the key part and the reference part, and a three-dimensional coordinate point corresponding to each pixel information;
acquiring a predetermined amount of plane feature point information for constructing a plane according to the extracted image information of the key part and a first extraction rule, wherein the plane feature point information comprises pixel information corresponding to the plane feature point and a three-dimensional coordinate point corresponding to the pixel information;
acquiring height reference characteristic point information according to the extracted image information of the reference part and a second extraction rule, wherein the height reference characteristic point information comprises pixel information corresponding to the height reference characteristic point and a three-dimensional coordinate point corresponding to the pixel information;
constructing a reference cube reflecting the back slimming of the animal according to the height reference feature point information, the plane feature point information and a preset height value, wherein the upper surface and the lower surface of the reference cube are the same in shape and size, the upper surface and the lower surface are parallel and opposite, and the height reference feature point is positioned in the upper surface of the reference cube;
calculating the volume of the reference cube according to the plane feature points and the preset height value, and taking the calculated volume value as a reference fat-lean volume value V whole ；
Calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube according to the pixel information in the target image area surrounded by the plane feature points, the three-dimensional coordinate point corresponding to the pixel information and the three-dimensional coordinate point of the lower surface of the reference cube;
calculating the volume value corresponding to each pixel in the target image region according to the relative height value, and integrating the volume of each pixel in the target image region to obtain the actual fat-lean volume value V reflecting animal fat and lean reality ；
Substituting the reference fat-lean volume value and the actual fat-lean volume value into a formula I for calculation so as to obtain a body condition score value BCS of the animal;
wherein δ in the formula one is a correction coefficient.
According to the animal body condition grading system and the animal body condition grading method, a back depth image reflecting the back contour of an animal is obtained from the upper part of the animal to be tested, image information reflecting animal fat and thin key parts and image information reflecting reference parts are extracted from the back depth image, a preset amount of plane feature point information used for constructing a plane is obtained according to the extracted image information of the key parts, a height reference feature point information is obtained according to the extracted image information of the reference parts, and a reference cube reflecting the animal back fat and thin is constructed according to the height reference feature point information, the plane feature point information and a preset height value to calculate the volume of the reference cube; according to pixel information in a target image area defined by the plane feature points, three-dimensional coordinate points corresponding to the pixel information and three-dimensional coordinate points of the lower surface of the reference cube, calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube, calculating a volume value corresponding to each pixel in the target image area according to the relative height value, integrating the volume of each pixel in the target image area to obtain an actual fat and thin volume reflecting the fat and thin of the animal, multiplying the actual fat and thin volume value and the reference fat and thin volume value by a correction coefficient to obtain a body condition evaluation value BCS of the animal, and further realizing body condition evaluation of the animal without manual observation and analysis.
Description of the drawings:
FIG. 1 is a functional block diagram of a preferred embodiment of a cattle body condition scoring system.
Fig. 2 is a functional block diagram of the body condition scoring module of fig. 1.
Fig. 3 is a schematic diagram of determination of points 1 to 7 of the back depth image of the cow.
Fig. 4 is a schematic diagram of the left side volume of fat, lean key parts of a cow.
Fig. 5 is a right side schematic volume of fat, lean key parts of a cow.
FIG. 6 is a flow chart of a method for scoring a body condition of an animal according to a preferred embodiment.
In the figure: the animal body condition scoring system 10, the depth image acquisition module 20, the body condition scoring module 30, the image information extraction unit 31, the first extraction unit 32, the second extraction unit 33, the reference fat-thin volume calculation unit 34, the actual fat-thin volume calculation unit 35, the body condition scoring calculation unit 36, the lookup information table generation unit 37, the body condition score value output unit 38, the animal identity recognition device 40, and the animal body condition scoring method steps S300 to S314.
The specific implementation mode is as follows:
the animal body condition scoring system provided by the invention obtains an ideal reference fat-lean volume value and an actual back volume value of the back of an animal by analyzing the back depth image of the animal, and obtains the body condition score value of the animal by multiplying the ratio of the actual fat-lean volume value to the reference fat-lean volume value by a coefficient so as to realize the scoring of the body condition of the animal.
Referring to fig. 1, the animal condition scoring system 10 includes a depth image obtaining module 20 and an animal condition scoring module 30; the depth image acquiring module 20 is used for acquiring a back depth image reflecting the back contour of the animal from above the animal to be tested, and providing the back depth image to the body condition scoring module 30. The depth image obtaining module 20 may be a depth camera. The depth image can accurately know the spatial distance of each pixel point in the image, and thus, the three-dimensional spatial coordinates of each point in the image can be obtained by adding the (x, y) coordinates of each pixel point in the two-dimensional image.
The body condition scoring module 30 is used for extracting image information representing animal fat and thin key parts and image information of a reference part from the acquired back depth image; the image information comprises pixel information reflecting the key part and the reference part, and a three-dimensional coordinate point corresponding to each pixel information; acquiring a predetermined amount of plane feature point information for constructing a plane according to the extracted image information of the key part and a first extraction rule, wherein the plane feature point information comprises pixel information corresponding to the plane feature point and a three-dimensional coordinate point corresponding to the pixel information; acquiring height reference characteristic point information according to the extracted image information of the reference part and a second extraction rule, wherein the height reference characteristic point information comprises pixel information corresponding to the height reference characteristic point and a three-dimensional coordinate point corresponding to the pixel information; constructing a reference cube reflecting the back slimming of the animal according to the height reference feature point information, the plane feature point information and a preset height value, wherein the upper surface and the lower surface of the reference cube are the same in shape and size, the upper surface and the lower surface are parallel and opposite, and the height reference feature point is positioned in the upper surface of the reference cube; for example, the animal is a cow, please see fig. 3 to 5 at the same time, the key parts are the hip angle, the tail root and the hip bone of the animal, in the hip angle image information, the two hip angle edges are marked as a point No. 4 and a point No. 6, and the point No. 4 endpoint pixel information and the point No. 6 endpoint pixel information are obtained from the hip angle image information according to the marked point No. 4 and point No. 6; marking the tail root edge as a No. 5 point in the tail root image information, and acquiring No. 5 point endpoint pixel information from the tail root image information according to the marked No. 5 point; marking a point No. 1 and a point No. 3 in two opposite edges in the hip bone image information, wherein the distance between the point No. 1 and the point No. 3 is longest, obtaining point No. 1 endpoint pixel information and point No. 3 endpoint pixel information from the hip bone image information according to the marked point No. 1 and point No. 3, determining point No. 2 endpoint pixel information in the back depth image according to the requirement of an equilateral triangle according to the length between the point No. 1 endpoint pixel information and the point No. 3 endpoint pixel information, and keeping the point No. 2 endpoint pixel information away from the point No. 5 endpoint pixel information; the above-mentioned number 1 to 6 end point pixel information is taken as plane feature point information. Marking a No. 7 point in the image information of the hip nodule, and obtaining No. 7 point endpoint pixel information from the image information of the hip nodule according to the marked No. 7 point, wherein the No. 7 point is the point with the highest distance from the ground in the back depth image, and the No. 7 point endpoint pixel information is used as height reference characteristic point information; determining a first plane (plane ABCDFE) with the height reference feature point information as a height reference point according to the plane feature point information and a preset height value (the preset height value is determined according to different animals, for example, in the embodiment, the animal is a cow, and the preset height value is 20cm), wherein the first plane has a lower mapping feature point corresponding to the plane feature point; and determining a second plane according to the plane feature point information and the height reference feature point information, wherein the second plane is provided with an upper mapping feature point corresponding to the plane feature point, the height reference feature point is positioned in the second plane, the first plane is parallel to the second plane, and the distance between the first plane and the second plane is the preset height value.
The body condition scoring module 30 calculates the volume of the reference cube according to the plane feature points and the preset height values, and uses the calculated volume value as a reference fat and thin volume value V whole (ii) a Calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube according to the pixel information in the target image area surrounded by the plane feature points, the three-dimensional coordinate point corresponding to the pixel information and the three-dimensional coordinate point of the lower surface of the reference cube; calculating the volume value corresponding to each pixel in the target image area according to the relative height value, and integrating the volume of each pixel in the target image area to obtain the actual fat and thin volume value V reflecting animal fat and thin reality Substituting the reference fat-lean volume value and the actual fat-lean volume value into a formula I for calculation so as to obtain a body condition score value BCS of the animal;
wherein, δ in the first formula is a correction coefficient, and δ is an empirical value determined by a worker according to actual production experience in actual production.
Further, the animal body condition scoring system 10 further comprises an animal identification device 40,
the animal identity recognition device 40 is used for acquiring identification information carried by an animal and providing the identification information to the body condition scoring module; the body condition scoring module 30 is further configured to establish a corresponding relationship according to the identification information and the back depth image, and form a lookup information table; the body condition scoring module 30 is further configured to display the obtained body condition scoring value BCS of the animal and the corresponding identification information according to the lookup information table.
The body condition scoring module 30 may be a single chip microcomputer or a microcomputer running a set of computer application programs, wherein the computer application programs are used for completing the body condition scoring function of the animals. After the computer application program is run, the body condition scoring module generates the following functional modules, please refer to fig. 2 at the same time:
the body condition scoring module 30 includes an image information extraction unit 31, a first extraction unit 32, a second extraction unit 33, a reference fat-lean volume calculation unit 34, an actual fat-lean volume calculation unit 35, a body condition scoring calculation unit 36, a lookup information table generation unit 37, and a body condition score value output unit 38;
the image information extraction unit 31 is used for extracting image information of key parts reflecting animal fatness and thinness and image information of reference parts from the acquired back depth image; the image information comprises pixel information reflecting the key part and the reference part and a three-dimensional coordinate point corresponding to the pixel information;
the first extraction unit 32 is configured to obtain, according to the extracted image information of the key portion, a predetermined number of pieces of plane feature point information for constructing a plane according to a first extraction rule, where the plane feature point information includes pixel information corresponding to the plane feature point and a three-dimensional coordinate point corresponding to the pixel information;
the second extraction unit 33 is configured to obtain height reference feature point information according to a second extraction rule based on the extracted image information of the reference portion, where the height reference feature point information includes pixel information corresponding to the height reference feature point and a three-dimensional coordinate point corresponding to the pixel information;
the standard slimming volume calculation unit 34 is configured to construct a standard cube reflecting the slimming of the back of the animal according to the height reference feature point information, the plane feature point information and the preset height value, wherein the upper surface and the lower surface of the standard cube have the same shape and size, the upper surface and the lower surface are parallel and opposite, and the height reference feature point is located in the upper surface of the standard cube; calculating the volume of the reference cube according to the plane feature points and the preset height value, and taking the calculated volume value as a reference fat-lean volume value V whole (ii) a For example, in fig. 4, 20cm is mapped downwards in the z-axis direction of the points 2, 3, 4 and 5 respectively to obtain corresponding points A, B, C, D, so that an irregular cube of the left side of the cow can be constructed, and the volume is defined as V 1 (ii) a ABCD dot composition V 1 Lower bottom surface S of 1 Then V 1 ＝S 1 20, x; FIG. 5 is a right side spatial view of a key point of a depth image of the back of a cow, and according to the key points No. 1, No. 2, No. 5 and No. 6, an irregular cube of the right side of the cow is constructed by the same method, and the volume is defined as V 2 AEFD four-point composition V 2 Lower bottom surface S of 2 ，V 2 ＝S 2 20, x; constructing an irregular cube on the back of the cow according to the key points 1, 2, 3, 4, 5 and 6 by the same method, wherein the volume is defined as V whole ＝V 1 +V 2 。
The actual fat and thin volume calculating unit 35 is configured to calculate a relative height value from each pixel located in the target image region to the lower surface of the reference cube according to the pixel information in the target image region surrounded by the planar feature points, the three-dimensional coordinate point corresponding to the pixel information, and the three-dimensional coordinate point of the lower surface of the reference cube; calculating the volume value corresponding to each pixel in the target image region according to the relative height value, and integrating the volume of each pixel in the target image region to obtain the actual fat-lean volume value V reflecting animal fat and lean reality (ii) a For example, a depth camera shoots a depth image of the back of a cow, RGBD data of a target pixel point is obtained, three-dimensional space coordinates of each point are known, real fat condition scenes of the back of different cows are restored, and the actual back fat condition of the cow is converted into an actual fat-lean volume value V reality To reflect the actual fat and lean volume value V reality The back of the corresponding cow is divided into a left side volume and a right side volume which are defined as V left And V right And using the concept of calculus to convert V left ，V right The volume is respectively assumed to be formed by n small cuboids, the length Ln, the width Wn and the height Hn of each small cuboid can be obtained by knowing the three-dimensional space coordinates of each point according to the RGBD data of the obtained target pixel points, the volume Vn can be calculated to be Ln Wn Hn, then the left side volume and the right side volume are obtained by using a formula II, and the left side volume and the right side volume are summed to obtain the whole actual fat and thin volume value V reality ，
The body condition score calculating unit 36 is used for substituting the reference fat-lean volume value and the actual fat-lean volume value into a prestored formula I for calculation so as to obtain a body condition score value BCS of the animal;
wherein, δ in the first formula is a correction coefficient, and δ is an empirical value determined by a worker according to actual production experience in actual production.
The lookup information table generating unit 37 is configured to establish a corresponding relationship according to the identification information and the back depth image, and form a lookup information table.
The body condition score value output unit 38 is configured to display or provide the obtained body condition score value BCS of the animal and the corresponding identification information to third-party application software according to the lookup information table, so as to display the animal condition score value BCS and the corresponding identification information by using the third-party application software. For example, the body condition score value output unit 38 may provide the body condition score value BCS of the animal and the corresponding identification information to the cloud platform, the mobile phone APP.
In other embodiments, the functional modules generated by the body condition scoring module can also be implemented by hardware logic circuits and embedded codes.
Further, the present invention also provides a method for scoring animal body conditions, please refer to fig. 6, which comprises the following steps:
and step S300, acquiring a back depth image reflecting the back contour of the animal from the upper part of the animal to be detected by using a depth camera.
Step S302, extracting image information of key parts reflecting animal fat and thinness and image information of reference parts from the acquired back depth image; the image information comprises pixel information reflecting the key part and the reference part, and a three-dimensional coordinate point corresponding to each pixel information;
step S304, according to the extracted image information of the key part, acquiring a predetermined number of plane feature point information for constructing a plane according to a first extraction rule, wherein the plane feature point information comprises pixel information corresponding to the plane feature point and a three-dimensional coordinate point corresponding to the pixel information. Wherein the first extraction rule is: and determining the endpoint pixel information of each key part according to the edge of each key part, and recording each endpoint pixel information as plane feature point information.
Step S306, according to the extracted image information of the reference part, acquiring height reference characteristic point information according to a second extraction rule, wherein the height reference characteristic point information comprises pixel information corresponding to the height reference characteristic point and a three-dimensional coordinate point corresponding to the pixel information; wherein the second extraction rule is: and selecting the highest pixel information from the pixel information of the reference position, and taking the selected highest pixel information as the height reference characteristic point information.
And step S308, constructing a reference cube reflecting the back slimming of the animal according to the height reference feature point information, the plane feature point information and a preset height value, wherein the upper surface and the lower surface of the reference cube are the same in shape and size, the upper surface and the lower surface are parallel and opposite, and the height reference feature point is positioned in the upper surface of the reference cube.
Step S310, calculating the volume of the reference cube according to the plane feature points and the preset height value, and taking the calculated volume value as a reference fat-lean volume value V whole 。
Step S312, calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube according to the pixel information in the target image area surrounded by the plane feature points, the three-dimensional coordinate point corresponding to the pixel information and the three-dimensional coordinate point of the lower surface of the reference cube, calculating a volume value corresponding to each pixel in the target image area according to the relative height value, and integrating the volume of each pixel in the target image area to obtain an actual fat-lean volume value V reflecting animal fat and lean reality 。
Step S314, substituting the reference fat-lean volume value and the actual fat-lean volume value into a formula I for calculation so as to obtain a body condition score value BCS of the animal;
wherein, δ in the formula one is a correction coefficient, and δ is an empirical value determined by a worker in actual production according to actual production experience.
Wherein, step S304 specifically includes: the key parts are the hip angle, the tail root and the hip bone of the animal, in the hip angle image information, two hip angle edges are marked as a No. 4 point and a No. 6 point, and No. 4 point end point pixel information and No. 6 point end point pixel information are obtained from the hip angle image information according to the marked No. 4 point and No. 6 point; marking the tail root edge as a No. 5 point in the tail root image information, and acquiring No. 5 point endpoint pixel information from the tail root image information according to the marked No. 5 point; marking a point No. 1 and a point No. 3 in two opposite edges in the hip bone image information, wherein the distance between the point No. 1 and the point No. 3 is longest, obtaining point No. 1 endpoint pixel information and point No. 3 endpoint pixel information from the hip bone image information according to the marked point No. 1 and point No. 3, determining point No. 2 endpoint pixel information in the back depth image according to the requirement of an equilateral triangle according to the length between the point No. 1 endpoint pixel information and the point No. 3 endpoint pixel information, and keeping the point No. 2 endpoint pixel information away from the point No. 5 endpoint pixel information; the above-mentioned number 1 to 6 end point pixel information is taken as plane feature point information.
Wherein, step S306 specifically includes: and marking a No. 7 point in the image information of the hip nodule, and acquiring No. 7 point endpoint pixel information from the image information of the hip nodule according to the marked No. 7 point, wherein the No. 7 point is the point with the highest distance from the ground in the back depth image, and the No. 7 point endpoint pixel information is used as height reference characteristic point information.
Wherein, step S312 specifically includes: determining a first plane according to the plane feature point information and a preset height value by taking the height reference feature point information as a height reference point, wherein the first plane is provided with a lower mapping feature point corresponding to the plane feature point; and determining a second plane according to the plane feature point information and the height reference feature point information, wherein the second plane is provided with an upper mapping feature point corresponding to the plane feature point, the height reference feature point is positioned in the second plane, the first plane is parallel to the second plane, and the distance between the first plane and the second plane is the preset height value.
Further, the animal body condition scoring method further comprises the following steps: acquiring identification information carried by an animal;
establishing a corresponding relation for the identification information and the back depth image;
and establishing a corresponding relation according to the identification information and the back depth image, and displaying the obtained body condition score value BCS of the animal and the corresponding identification information.
In the animal body condition evaluation system 10 and method, a back depth image reflecting the back contour of an animal is obtained from the upper part of the animal to be tested, image information representing the fat and thin key parts of the animal and image information of a reference part are extracted from the back depth image, a predetermined number of plane feature point information for constructing a plane is obtained according to the extracted image information of the key parts, a height reference feature point information is obtained according to the extracted image information of the reference part, and a reference cube reflecting the fat and thin back of the animal is constructed according to the height reference feature point information, the plane feature point information and a preset height value to calculate the volume of the reference cube; according to pixel information in a target image area surrounded by the plane feature points, three-dimensional coordinate points corresponding to the pixel information and three-dimensional coordinate points of the lower surface of the reference cube, calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube, calculating a volume value corresponding to each pixel in the target image area according to the relative height value, integrating the volume of each pixel in the target image area to obtain an actual fat and thin volume reflecting the fat and thin of the animal, multiplying the actual fat and thin volume value and the reference fat and thin volume value by a correction coefficient delta to obtain a body condition score value BCS of the animal, and further realizing body condition scoring of the animal without manual observation and analysis. δ is an empirical value determined by a worker in real production based on actual production experience, for example, δ is equal to 5 points mentioned in the background of the present application, and δ may be equal to 5 points multiplied by a conventional coefficient in other embodiments. Animal body condition scores are measures of the condition of the animal's fat, which reflect the basic condition of fat deposition in the animal. By knowing the body condition scores of the groups and individuals, the feeding effect in the period can be researched and evaluated, and important basis is provided for feeding measures in the next stage, and adjustment of the recent daily ration formula and the feeding amount. In addition, body condition scoring is also an aid to animal health checks.
Claims (10)
1. An animal condition scoring system, comprising: the system comprises a depth image acquisition module and a body condition scoring module;
the depth image acquisition module is used for acquiring a back depth image reflecting the back contour of the animal from the upper part of the animal to be detected and providing the back depth image for the body condition scoring module;
the body condition scoring module is used for extracting body motion from the acquired back depth imageThe image information of the key part of the fertilizer and the thin part and the image information of the reference part; the image information comprises pixel information reflecting the key part and the reference part, and a three-dimensional coordinate point corresponding to each pixel information; acquiring a predetermined amount of plane feature point information for constructing a plane according to the extracted image information of the key part and a first extraction rule, wherein the plane feature point information comprises pixel information corresponding to the plane feature point and a three-dimensional coordinate point corresponding to the pixel information; acquiring height reference characteristic point information according to the extracted image information of the reference part and a second extraction rule, wherein the height reference characteristic point information comprises pixel information corresponding to the height reference characteristic point and a three-dimensional coordinate point corresponding to the pixel information; constructing a reference cube reflecting the back slimming of the animal according to the height reference feature point information, the plane feature point information and a preset height value, wherein the upper surface and the lower surface of the reference cube are the same in shape and size, the upper surface and the lower surface are parallel and opposite, and the height reference feature point is positioned in the upper surface of the reference cube; calculating the volume of the reference cube according to the plane feature points and the preset height value, and taking the calculated volume value as a reference fat-lean volume value V whole (ii) a Calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube according to the pixel information in the target image area surrounded by the plane feature points, the three-dimensional coordinate point corresponding to the pixel information and the three-dimensional coordinate point of the lower surface of the reference cube; calculating the volume value corresponding to each pixel in the target image region according to the relative height value, and integrating the volume of each pixel in the target image region to obtain the actual fat-lean volume value V reflecting animal fat and lean reality (ii) a Substituting the reference fat-lean volume value and the actual fat-lean volume value into a formula I for calculation so as to obtain a body condition score value BCS of the animal;
wherein δ in the first formula is a correction coefficient.
2. An animal condition scoring system according to claim 1, wherein: the body condition grading module comprises an image information extraction unit, a first extraction unit, a second extraction unit, a reference fat and thin volume calculation unit, an actual fat and thin volume calculation unit and a body condition grading calculation unit;
the image information extraction unit is used for extracting image information of key parts reflecting animal fatness and thinness and image information of reference parts from the acquired back depth image; the image information comprises pixel information reflecting the key part and the reference part and a three-dimensional coordinate point corresponding to the pixel information;
the first extraction unit is used for acquiring a predetermined number of pieces of plane characteristic point information for constructing a plane according to the extracted image information of the key part and a first extraction rule, wherein the plane characteristic point information comprises pixel information corresponding to the plane characteristic points and three-dimensional coordinate points corresponding to the pixel information;
a second extraction unit, configured to obtain height reference feature point information according to a second extraction rule based on the extracted image information of the reference portion, where the height reference feature point information includes pixel information corresponding to the height reference feature point and a three-dimensional coordinate point corresponding to the pixel information;
the standard slimming volume calculation unit is used for constructing a standard cube reflecting the slimming back of the animal according to the height reference characteristic point information, the plane characteristic point information and a preset height value, wherein the shape and the size of the upper surface and the lower surface of the standard cube are the same, the upper surface and the lower surface are parallel and opposite, and the height reference characteristic point is positioned in the upper surface of the standard cube; calculating the volume of the reference cube according to the plane feature points and the preset height value, and taking the calculated volume value as a reference fat-lean volume value V whole ；
The actual fat and thin volume calculation unit is used for calculating the pixel information in a target image region surrounded by the plane characteristic points and the three-dimensional coordinate point and the base corresponding to the pixel informationCalculating a relative height value from each pixel in the target image area to the lower surface of the reference cube by using the three-dimensional coordinate points of the lower surface of the quasi-cube; calculating the volume value corresponding to each pixel in the target image region according to the relative height value, and integrating the volume of each pixel in the target image region to obtain the actual fat-lean volume value V reflecting animal fat and lean reality ；
The body condition scoring calculation unit is used for substituting the reference fat-lean volume value and the actual fat-lean volume value into a prestored formula I for calculation so as to obtain a body condition scoring value BCS of the animal;
wherein δ in the formula one is a correction coefficient.
3. An animal condition scoring system as claimed in claim 1 or 2, wherein: the animal body condition scoring system also comprises an animal identity recognition device,
the animal identity recognition device is used for acquiring identification information carried by an animal and providing the identification information to the body condition scoring module;
the body condition scoring module is also used for establishing a corresponding relation according to the identification information and the back depth image to form a lookup information table; and the body condition scoring module is also used for displaying the obtained body condition scoring value BCS of the animal and the corresponding identification information according to the lookup information table.
4. An animal condition scoring system as claimed in claim 3, wherein: the first extraction rule is as follows: determining the endpoint pixel information of each key part according to the edge of each key part, and recording the endpoint pixel information as plane feature point information; the second extraction rule is: and selecting the highest pixel information from the pixel information of the reference position, and taking the selected highest pixel information as the height reference characteristic point information.
5. A method of scoring a body condition of an animal comprising the steps of:
acquiring a back depth image reflecting the back contour of an animal from the upper part of the animal to be detected by using a depth camera;
extracting image information of key parts reflecting animal fatness and thinness and image information of reference parts from the acquired back depth image; the image information comprises pixel information reflecting the key part and the reference part, and a three-dimensional coordinate point corresponding to each pixel information;
acquiring a predetermined amount of plane feature point information for constructing a plane according to the extracted image information of the key part and a first extraction rule, wherein the plane feature point information comprises pixel information corresponding to the plane feature point and a three-dimensional coordinate point corresponding to the pixel information;
acquiring height reference characteristic point information according to the extracted image information of the reference part and a second extraction rule, wherein the height reference characteristic point information comprises pixel information corresponding to the height reference characteristic point and a three-dimensional coordinate point corresponding to the pixel information;
constructing a reference cube reflecting the back slimming of the animal according to the height reference feature point information, the plane feature point information and a preset height value, wherein the upper surface and the lower surface of the reference cube are the same in shape and size, the upper surface and the lower surface are parallel and opposite, and the height reference feature point is positioned in the upper surface of the reference cube;
calculating the volume of the reference cube according to the plane feature points and the preset height value, and taking the calculated volume value as a reference fat-lean volume value V whole ；
Calculating a relative height value from each pixel in the target image area to the lower surface of the reference cube according to the pixel information in the target image area surrounded by the plane feature points, the three-dimensional coordinate point corresponding to the pixel information and the three-dimensional coordinate point of the lower surface of the reference cube;
calculating the corresponding of each pixel in the target image area according to the relative height valueIntegrating the volume of each pixel in the target image region to obtain actual fat-lean volume value V reflecting animal fat and lean reality ；
Substituting the reference fat-lean volume value and the actual fat-lean volume value into a formula I to calculate so as to obtain a body condition score value BCS of the animal;
wherein δ in the formula one is a correction coefficient.
6. The animal condition scoring method of claim 5, wherein the first extraction rule is: determining the endpoint pixel information of each key part according to the edge of each key part, and recording the endpoint pixel information as plane feature point information; the second extraction rule is as follows: and selecting the highest pixel information from the pixel information of the reference position, and taking the selected highest pixel information as the height reference characteristic point information.
7. The animal body condition scoring method according to claim 6, wherein the step of "obtaining a predetermined number of pieces of plane feature point information for constructing a plane according to the first extraction rule based on the image information of the extracted key parts, wherein the plane feature point information includes pixel information corresponding to the plane feature points and three-dimensional coordinate points corresponding to the pixel information" is specifically:
the key parts are the hip angle, the tail root and the hip bone of the animal, in the hip angle image information, two hip angle edges are marked as a No. 4 point and a No. 6 point, and No. 4 point end point pixel information and No. 6 end point pixel information are obtained from the hip angle image information according to the marked No. 4 point and No. 6 point; marking the tail root edge as a No. 5 point in the tail root image information, and acquiring No. 5 point endpoint pixel information from the tail root image information according to the marked No. 5 point; marking a point No. 1 and a point No. 3 in two opposite edges in the hip bone image information, wherein the distance between the point No. 1 and the point No. 3 is longest, obtaining point No. 1 endpoint pixel information and point No. 3 endpoint pixel information from the hip bone image information according to the marked point No. 1 and point No. 3, determining point No. 2 endpoint pixel information in the back depth image according to the requirement of an equilateral triangle according to the length between the point No. 1 endpoint pixel information and the point No. 3 endpoint pixel information, and keeping the point No. 2 endpoint pixel information away from the point No. 5 endpoint pixel information; the above-mentioned number 1 to 6 end point pixel information is taken as plane feature point information.
8. The animal body condition scoring method according to claim 6 or 5, wherein the step of "obtaining, based on the extracted image information of the reference portion, a height reference feature point information according to a second extraction rule, the height reference feature point information including pixel information corresponding to the height reference feature point and a three-dimensional coordinate point corresponding to the pixel information" is embodied as: and marking a No. 7 point in the image information of the hip nodule, and acquiring No. 7 point endpoint pixel information from the image information of the hip nodule according to the marked No. 7 point, wherein the No. 7 point is the point with the highest distance from the ground in the back depth image, and the No. 7 point endpoint pixel information is used as height reference characteristic point information.
9. The animal body condition scoring method according to claim 8, wherein the step of constructing a reference cube reflecting the back slimming of the animal based on the height reference feature point information, the plane feature point information, and the preset height value, wherein the upper surface and the lower surface of the reference cube have the same shape and size, are parallel to and opposite to each other, and the height reference feature point is located in the upper surface of the reference cube, specifically comprises: determining a first plane according to the plane feature point information and a preset height value by taking the height reference feature point information as a height reference point, wherein the first plane is provided with a lower mapping feature point corresponding to the plane feature point; and determining a second plane according to the plane feature point information and the height reference feature point information, wherein the second plane is provided with an upper mapping feature point corresponding to the plane feature point, the height reference feature point is positioned in the second plane, the first plane is parallel to the second plane, and the distance between the first plane and the second plane is the preset height value.
10. A method of scoring a condition of an animal as claimed in claim 9, further comprising the steps of: acquiring identification information carried by an animal;
establishing a corresponding relation for the identification information and the back depth image;
and establishing a corresponding relation according to the identification information and the back depth image, and displaying the obtained body condition score value BCS of the animal and the corresponding identification information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210361831.6A CN114821400A (en) | 2022-04-07 | 2022-04-07 | Animal body condition evaluation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210361831.6A CN114821400A (en) | 2022-04-07 | 2022-04-07 | Animal body condition evaluation system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114821400A true CN114821400A (en) | 2022-07-29 |
Family
ID=82534223
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210361831.6A Pending CN114821400A (en) | 2022-04-07 | 2022-04-07 | Animal body condition evaluation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114821400A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116363141A (en) * | 2023-06-02 | 2023-06-30 | 四川省畜牧科学研究院 | Pregnant sow intelligent body type evaluation device and system |
-
2022
- 2022-04-07 CN CN202210361831.6A patent/CN114821400A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116363141A (en) * | 2023-06-02 | 2023-06-30 | 四川省畜牧科学研究院 | Pregnant sow intelligent body type evaluation device and system |
| CN116363141B (en) * | 2023-06-02 | 2023-08-18 | 四川省畜牧科学研究院 | Pregnant sow intelligent body type evaluation device and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5519693B2 (en) | Apparatus and method for determining a score of an animal's physical condition | |
| CN107180438B (en) | Method for estimating size and weight of yak and corresponding portable computer device | |
| CN110426112B (en) | Live pig weight measuring method and device | |
| Wu et al. | Extracting the three-dimensional shape of live pigs using stereo photogrammetry | |
| CN111243005B (en) | Livestock weight estimation method, apparatus, device and computer readable storage medium | |
| JP2021515323A (en) | Methods and devices for remotely characterizing biological specimens | |
| Gaber et al. | Quantifying facial paralysis using the kinect v2 | |
| CN110782467B (en) | Horse body ruler measuring method based on deep learning and image processing | |
| CN113222991A (en) | Deep learning network-based field ear counting and wheat yield prediction | |
| JP7057971B2 (en) | Animal body weight estimation device and weight estimation method | |
| CN114821400A (en) | Animal body condition evaluation system and method | |
| CN110530477A (en) | Replacement gilt weight predictor method | |
| CN112288793A (en) | Livestock individual backfat detection method and device, electronic equipment and storage medium | |
| CN106023235B (en) | A kind of method that the effective seed number of crops measures | |
| CN108830154A (en) | A kind of food nourishment composition detection method and system based on binocular camera | |
| CN115969405A (en) | Oral implantation precision evaluation method based on CT image | |
| CN106767433A (en) | A kind of method and system for measuring foot sizing | |
| CN110414369B (en) | Cow face training method and device | |
| CN112258538A (en) | Method and device for acquiring three-dimensional data of human body | |
| CN114821637A (en) | Animal gait scoring system and method | |
| Kawasue et al. | Pig weight prediction system using RGB-D sensor and AR glasses: analysis method with free camera capture direction | |
| CN112598079A (en) | Method for identifying cephalopod population and species | |
| McFarlane et al. | Shape measurements of live pigs using 3-D image capture | |
| CN116309641B (en) | Image area acquisition system | |
| Siachos et al. | Development and validation of a fully automated 2-dimensional imaging system generating body condition scores for dairy cows using machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |