Mark application system, mark application program, and mark application method

Abstract

It allows for the precise application of marks to target animals of the desired weight without requiring manpower or time. [Solution] The system works in cooperation with a weight management and holding system 200 that obtains image information of animals from a weight measurement camera, performs individual identification of animals based on the image information, measures the weight of the animals, and manages and holds set information which is a set of individual identification information, weight information and image information used for individual identification. The system includes a mark attachment area camera 160 for obtaining image information of animals in the mark attachment area, and a mark attachment mechanism 170 for attaching marks to animals in the mark attachment area.

Description

Technical Field

[0001] The present invention relates to a marking system, a marking program, and a marking method for attaching a mark to an animal whose weight has reached a predetermined value.

Background Art

[0002] Regarding animal weight measurement, those described in the following Patent Documents 1 to 3 are known. The weight measuring device for pigs described in Patent Document 1 has a dust-proof box mounted on the ceiling of a piggery near a water feeder, and a projector equipped with a checkered slide and a video camera are installed in the dust-proof box. When light is projected from the projector, checkered patterns are projected onto the floor and the body surface of the pig. This image is captured by a video camera, and the body height (H) of the pig is obtained based on the deviation of the checkered patterns in the captured image. Further, black-and-white binarization processing is performed on the captured image to obtain the projected area (A) of the pig. There is a close relationship expressed as W=aHbAc among the pig's body height (H), projected area (A), and body weight (W). Furthermore, since the dimension (d=b+2c) of the multiple regression equation is close to the dimension of volume (d=3), the body weight (W) is calculated using the body height (H) and the projected area (A).

[0003] Patent Document 2 discloses that, in the fish culture process or at the time of shipment, three types of sorting criteria are provided: only body length, both body length and weight, and only weight; by selecting the criteria, fish can be efficiently and accurately measured for body length and weight without weakening the live fish, and a measuring device capable of sorting in multiple stages is provided.

[0004] This measuring device includes a weighing chute that slopes downwards and flows a constant amount of water, and a passage detector is mounted in the middle of the weighing chute to detect fish passing through it. A light source and camera capable of capturing still images of the fish based on the detection signal from this detector are mounted above and below the weighing chute. Meanwhile, the movable part of a load cell that emits an output proportional to the fish's weight is mounted on the weighing chute, and the fixed part is mounted on the weighing instrument frame. Image processing is performed by an image processing device and image memory built into the control device, and this image and the load cell output are used by the CPU in the control device to calculate body length and weight data, respectively. The sorting criterion setting device of the control device is configured to compare the input body length only, body length and weight, and weight only with three sorting criteria and sort the fish in multiple stages using a sorting device located downstream.

[0005] Patent Document 3 describes the provision of an animal weight measuring device, a measuring facility, and an automatic measuring device.

[0006] This weight measuring device consists of a load sensor, a measuring platform mounted horizontally above it and having a predetermined size according to the target animal, side fences positioned approximately vertically at the left and right ends of the measuring platform and having a predetermined height and depth according to the target animal, and a front end fence installed on the front end of the measuring platform and lower in height than the side fences. The rear end between the side fences is left open. The measurement facility consists of one or more weight measuring devices arranged in a ring or horizontal row, a feeder 11 positioned near the front of the front end fence, and a barrier member to prevent entry to the feeder side positioned outside the side fences. The automatic measurement device consists of a weight measuring device, a cable for transmitting its sensor signal, and a data acquisition and processing device. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Japanese Patent Publication No. 2002-286421 [Patent Document 2] Japanese Patent Application Publication No. 08-050052 [Patent Document 3] Japanese Patent Publication No. 2003-114145 [Overview of the project] [Problems that the invention aims to solve]

[0008] In the method described in Patent Document 1, it is difficult to project a grid pattern onto the floor and the surface of the pig's body, and accurate measurements are not always possible. In Patent Document 2, the method is based on measuring fish and has the problem of being difficult to apply to measuring the weight of other animals.

[0009] Patent Document 3 requires guiding the animals to the weighing platform, which is difficult. The biggest challenge for pig farms is the aging workforce and the declining working population. Weighing pigs requires a great deal of effort, which means that they cannot weigh the pigs and are not able to ship them at the optimal time.

[0010] As shown in Figure 1, pork is graded as standard, medium, high, and premium based on carcass weight and back fat height, and it is desirable to ship premium quality pork. Ideally, the carcass should be of high quality or high grade as shown in Figure 1, and the carcass weight should be large. To achieve this, weight management is basically carried out. Figure 2 shows the growth curve of a typical pig, and as an example, it is considered desirable to ship pigs weighing 120 kg or more.

[0011] Conventionally, as shown in Figure 3, a farmer P would enter the pig pen and visually select pigs weighing, for example, 120 kg or more, and mark them in the required location using marking spray M. Furthermore, as shown in Figure 4, multiple farmers P would guide the marked pigs to a weighing scale called a pig weighing machine for weight measurement. Based on the weight measurement, pigs that had reached the target weight were guided to shipping containers, while pigs that had not reached the target weight were returned to the pig pen. This process required about three farmers P and was time-consuming and labor-intensive overall.

[0012] The embodiments of the invention have been made in view of the conventional marking work on animals described above, and aim to provide a marking system, a marking program, and a marking method that can appropriately apply marks to target animals that have reached a desired weight without requiring manual labor or time. [Means for solving the problem]

[0013] The marking system according to this embodiment is characterized by comprising: an acquisition means for acquiring individual identification information and weight information of an animal related to the image information currently obtained by the marking area camera, using the image information of the animal obtained by the marking area camera and the set information managed and maintained by the weight management and maintenance system, which acquires image information of an animal from a weight measurement camera, performs individual identification of the animal based on the image information and measures the weight of the animal, and cooperates with a weight management and maintenance system which manages and maintains set information which is a set of individual identification information, weight information and image information used for individual identification, and for controlling the marking mechanism to execute or not execute the execution of

[0014] The marking program according to this embodiment works in cooperation with a weight management and storage system that obtains image information of animals from a weight measurement camera, performs individual identification of animals based on the image information, measures the weight of the animals, and manages and stores set information which is a set of individual identification information, weight information and image information used for individual identification. The program comprises a marking area camera for obtaining image information of animals in a marking area, and a marking mechanism for attaching marks to animals in the marking area. The computer provided in the marking system for attaching marks to animals in the marking area functions as an acquisition means for obtaining individual identification information and weight information of animals related to the image information currently obtained by the marking area camera using the image information of animals obtained by the marking area camera and the set information managed and stored in the weight management and storage system; a determination means that has weight threshold information for determining whether or not to attach a mark and determines whether or not to attach a mark using the weight information obtained by the acquisition means; and a control means that controls the marking mechanism based on the determination result of the determination means to execute or not execute the attachment of marks to animals in the marking area.

[0015] The mark application method according to this embodiment obtains image information of animals from a weight measurement camera, performs individual identification of animals based on the image information, measures the weight of the animals, and works in cooperation with a weight management and storage system that manages and stores set information consisting of individual identification information, weight information, and the image information used for individual identification. The system comprises a mark application area camera for obtaining image information of animals in the mark application area, and a mark application mechanism for applying marks to animals in the mark application area. The CPU used in the mark application system appropriately reads programs and data for operation from storage into main memory. A mark application method, which is performed by the CPU processing data, is characterized by comprising: an acquisition step of acquiring individual identification information and weight information of an animal related to the image information currently obtained by the mark application area camera, using the image information of the animal obtained by the mark application area camera and the set information managed and maintained by the weight management and holding system; a determination step of determining whether or not to apply a mark using the weight information obtained by the acquisition means, which has weight threshold information of whether or not to apply a mark; and a control step of controlling the mark application mechanism based on the determination result of the determination step to perform or not perform the application of a mark to an animal in the mark application area. [Brief explanation of the drawing]

[0016] [Figure 1] A diagram showing the weight of the pork carcass, the height of the back fat, and the resulting grade. [Figure 2] A diagram showing the growth curve of a typical pig. [Figure 3] A diagram illustrating the conventional marking process for pigs. [Figure 4] A diagram illustrating the conventional shipping process for pigs. [Figure 5] A schematic diagram of a mark application system according to an embodiment of the invention. [Figure 6] A diagram illustrating the configuration of a collaborative weight management and retention system configured using a computer in an embodiment of the present invention. [Figure 7] FIG. 1 is a functional block diagram of a weight measurement system according to a first embodiment of a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 8] FIG. 2 is an explanatory diagram of planar dimension and area information acquired by a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 9] FIG. 3 is an explanatory diagram of side dimension and area information acquired by a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 10] FIG. 4 is a flowchart showing the operation of the first embodiment of a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 11] FIG. 5 is a functional block diagram of a second embodiment of a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 12] FIG. 6 is an explanatory diagram of appropriate image selection performed by a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 13] FIG. 7 is an explanatory diagram of correction processing performed by a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 14] FIG. 8 is a flowchart showing the operation of the second embodiment of a collaborative weight management and maintenance system in an embodiment of the present invention. [Figure 15] FIG. 9 is a diagram showing estimation accuracy of a collaborative weight management and maintenance system for a certain period in an embodiment of the present invention. [Figure 16] FIG. 10 is a system configuration diagram drawn centered on an edge computer that constitutes a mark attachment system according to an embodiment of the present invention. [Figure 17] FIG. 11 is a functional block diagram of a mark attachment system according to an embodiment of the present invention. [Figure 18] FIG. 12 is a perspective view of an inverted conical feeder used in a mark attachment system according to an embodiment of the present invention. [Figure 19] FIG. 13 is a plan view showing a state where an animal is eating in the inverted conical feeder used in the mark attachment system according to an embodiment of the present invention. [Figure 20] FIG. 14 is a perspective view of a rectangular feeder used in a mark attachment system according to an embodiment of the present invention. [Figure 21] A plan view of an animal eating feed in a rectangular feeding machine used in a mark-attaching system according to an embodiment of the present invention. [Figure 22] A perspective view of an inverted cone-shaped feeding machine used in a marking system according to an embodiment of the present invention, with a marking attachment installed. [Figure 23] This figure shows an example of image information when detecting a mark of a predetermined color and area in a mark application system according to an embodiment of the present invention. [Figure 24] A flowchart illustrating the operation of a mark application system according to an embodiment of the present invention. [Figure 25] An explanatory diagram of the image information and set information obtained in the mark application system according to an embodiment of the present invention. [Modes for carrying out the invention]

[0017] The mark application system, mark application program, and mark application method according to the present invention will be described below with reference to the attached drawings. In each figure, the same components are denoted by the same reference numerals, and redundant explanations are omitted. Figure 5 shows a schematic configuration diagram of the mark application system 100 according to the embodiment of the invention. The mark application system 100 is a system that is connected to and cooperates with the weight management and holding system 200. The mark application system 100 comprises an edge computer 150, a mark application area camera 160, and a mark application mechanism 170. The weight management and holding system 200 comprises an edge computer 250 and a weight measurement camera 260.

[0018] The weight management and retention system 200 obtains image information of animals from a weight measurement camera 260, performs individual animal identification based on this image information, measures the weight of the animal, and manages and retains set information that includes the individual identification information, weight information, and the image information used for individual identification. Figure 6 shows a system configuration diagram centered on the edge computer 250 that constitutes the weight management and retention system 200. Specifically, the CPU 10 configures the weight management and retention system 200 using programs and data in the main memory 11. The CPU 10 is connected to a storage interface 13, an input interface 14, a display interface 15, and a data input interface 16 via a bus 12.

[0019] Storage interface 13 is connected to storage 23. Storage 23 stores programs and data necessary for the operation of this weight management and holding system 200, which the CPU 10 can read and use from main memory 11 as needed. Storage 23 includes external storage devices such as HDDs, auxiliary storage devices, and cloud storage. This storage 23 stores programs that implement machine learning models and various means described later. Input interface 14 is connected to input devices 24 such as keyboards and touch panels and pointing devices 22 such as mice. Display interface 15 is connected to a display device 25 having a screen such as an LCD. Cameras 26-1 to 26-m are connected to data input interface 16, and these cameras 26-1 to 26-m function to image the target animal. The number of cameras 26-1 to 26-m is arbitrary. Image data obtained by cameras 26-1 to 26-m is taken in by the data input interface 16 to the CPU 10, display interface 15, and other necessary parts. Cameras 26-1 to 26-m, specifically camera 260, are weight measurement cameras.

[0020] Figure 7 shows a functional block diagram of a weight management and holding system 200 according to the first embodiment of the present invention. In this embodiment, the system includes a camera 26 (26-1 to 26-m), an image acquisition means 31, an individual information management means 32, a planar dimension / area information acquisition means 33, a side dimension / area information acquisition means 34, a weight information acquisition means 40, and a calibration unit 60. These means can be implemented by the CPU 10 appropriately reading programs from the storage 23 into the main memory 11, or by using programs that are initially stored in the main memory 11. Furthermore, the calibration unit 60 may be implemented by the CPU 10 appropriately reading programs from the storage 23 into the main memory 11, but it may also be implemented by a cloud computer connected to the edge computer in Figure 6 via a network. The calibration unit 60 is also provided in the weight management and holding system 200 according to the second embodiment of the present invention, and will be described in detail after the description of the weight management and holding system 200 according to the second embodiment of the present invention.

[0021] Camera 26 (26-1 to 26-m) is a weight measurement camera 260, and camera 26 can also be a 3D camera. Image acquisition means 31 captures images of the target animal using camera 26 and obtains a planar image from above and a lateral image from the side of the target animal. Individual information management means 32 uses the images obtained by the image acquisition means 31 to perform individual identification of the target animal, issues and attaches identification information to the image of each individual target animal, and then provides the image with the attached identification information to the planar dimension / area information acquisition means 33, the lateral dimension / area information acquisition means 34, and the weight information acquisition means 40.

[0022] The plane dimension / area information acquisition means 33 obtains plane dimension / area information of the target animal based on the plane image obtained by the image acquisition means. In this embodiment, plane dimension / area information refers to plane width information and plane length information, but of course, dimension information in diagonal directions on the plane or area information of a predetermined part of the plane may also be used. The plane width information of the target animal will be explained. Assuming that the image of the target animal from above (here, an image facing left) is shown as U in Figure 8, the width of the animal image (length in the vertical direction in the figure) is measured from the center to the left in the area obtained by dividing this image vertically in the figure, and the measurement value of the largest divided area is taken as the shoulder width (Figure 8(B)). Similarly, the measurement is taken from the center to the right, and the measurement value of the largest divided area is taken as the hip width (Figure 8(B)). These can be used as plane width information. In addition, hip width information, etc., defined by a predetermined definition may also be used as plane width information. The planar width information can be at least one of the following: the shoulder width information of the target animal, the waist width information of the target animal, or the hip width information of the target animal.

[0023] Next, we will explain the planar length information. As described above, the positions of the shoulders and hips are determined in the process of obtaining the planar width information, so the distance between the shoulders and hips is taken as the body length. This body length can be used as the planar length information. The distance between the shoulders and waist can also be taken as the body length.

[0024] The lateral dimension / width information acquisition means 34 obtains lateral dimension / width information of the target animal based on the lateral image obtained by the image acquisition means 31. In this embodiment, the lateral dimension / width information is used as body height information, but of course, dimensional information in diagonal directions on the side or width information of a predetermined part of the side may also be used. The planar lateral dimension / width information of the target animal can be body height information, which is the body height information of at least one position of the target animal, such as the shoulder position, waist position, or rump position. For example, as shown in Figure 9, using a planar image U and a lateral image S of the target animal, first, the highest point in the same divided area of ​​the lateral image S is determined in the length direction. Measurements are taken from the center of the image to the left, and the height measurement of the divided area with the maximum is taken as the shoulder height. Measurements are taken from the center of the image to the right, and the height measurement of the divided area with the maximum is taken as the rump height. Alternatively, lateral dimension / width information of the waist, etc., defined by a predetermined definition, may be used as planar width information. The planar side dimension and width information can be body height information, which is body height information at at least one of the shoulder, waist, and rump positions of the target animal.

[0025] The weight information acquisition means 40 includes a machine learning model 41 obtained by machine learning, with the above-mentioned planar dimensions / area information and the above-mentioned side dimensions / area information as explanatory variables and the weight information of the target animal as the objective variable. This machine learning model 41 can be created by performing machine learning with the planar dimensions / area information and the above-mentioned side dimensions / area information obtained in this embodiment as explanatory variables and the weight information of the target animal measured as the objective variable. The weight information acquisition means 40 provides the obtained planar dimensions / area information and the above-mentioned side dimensions / area information as explanatory variables to the machine learning model 41 to obtain the weight information, which is the objective variable.

[0026] The weight management and retention system 200, configured as described above, performs operations according to a program corresponding to the flowchart shown in Figure 10. The operation will be described below based on the flowchart. The camera 26 captures an image of the target animal (S11). The process of capturing an image of the target animal with the camera 26 continues in the subsequent steps.

[0027] When the camera 26 captures an image of the target animal, the CPU 10 receives the captured image data (S12) and calculates planar width information, planar length information, and side dimensions / width information (S13). The CPU 10 provides the calculated planar width information, planar length information, and side dimensions / width information to a machine learning model to obtain the weight of the target animal (S14). In this way, it becomes possible to accurately measure the weight of the target animal. Thus, the weight of the target animal is measured, and a set of information consisting of individual identification information, weight information, and image information used for individual identification is stored in the storage 23.

[0028] Next, a weight management and holding system 200 according to a second embodiment will be described. Figure 11 shows a functional block diagram of the weight management and holding system 200 according to a second embodiment of the present invention. In this embodiment, the camera 26 and image acquisition means 31 are the same as in the first embodiment. In this embodiment, an individual information management means 352, an individual target animal image acquisition means 365, an appropriate image selection means 37, and a correction processing means 38 are provided. In this embodiment, the target animal is described as a pig, but the target animal of the present invention is not limited to this.

[0029] The individual information management means 352 uses the images obtained by the image acquisition means 31 to perform individual identification (image recognition) of the target animals, attaches identification information to the image of each individual target animal, and provides the images with this identification information to the planar dimension / area information acquisition means 33, the side dimension / area information acquisition means 34, and the weight information acquisition means 40. In other words, when multiple target animals are being raised, identification information is attached to each target animal to manage their images. More specifically, tracking is performed using RGB images, and a banding box is generated for each pig. If there are multiple pigs, a separate banding box and its coordinates are calculated for each pig, and these are passed to the individual target animal image acquisition means 365, which is the target selection unit.

[0030] This embodiment addresses cases where multiple target animals are kept together, and the camera 26 captures images of the area where multiple target animals are present. As a result, multiple target animals appear to overlap in the image. The individual target animal image acquisition means 365 separates the images of individual target animals from the images obtained by the camera 26 and acquires individual target animal images. In this embodiment, a segmentation method can be employed. By using this segmentation method, even overlapping pigs can be successfully separated, and mask images of each individual pig are generated and used for processing by the weight information acquisition means 40.

[0031] The appropriate image selection means 37 performs the process of selecting only images that can be properly processed when images with the above-mentioned identification information are provided to the planar dimension / area information acquisition means 33, the side dimension / area information acquisition means 34, and the weight information acquisition means 40. If the pig's posture is bent, the values ​​for each part cannot be measured correctly. Therefore, we considered what to do with images in a bent state. In this embodiment, since a fixed-point camera is used, the same pig can be photographed multiple times, so we decided to discard data in a bent state and use only those photographed in a good state.

[0032] Therefore, we decided to exclude images where the back is curved. Next, the determination of whether the body is curved can be made as follows. When obtaining planar dimensions and area information, multiple divided areas are created as shown in Figure 5, and the line connecting the highest points in the height direction of each divided area in Figure 12 corresponds to the spine. The posture of the pig can be determined by detecting the degree of curvature of this spine. Since an ellipse has been fitted in the prior image processing, the center line of this ellipse is the ideal position of the spine. Images with a large deviation from this line are judged to have high curvature and can be excluded (Figure 12(B)), while those with a small deviation from this line can be kept (Figure 12(A)). For example, images where the distance from the line connecting the highest points in the height direction of each divided area is greater than a predetermined value can be excluded. Also, since even pigs judged to be normal have a slight curve, deviation data from the ideal position indicating this spinal curvature (e.g., variance) can also be added as an explanatory variable.

[0033] The correction processing means 38 performs a correction process on the image selected by the appropriate image selection means 37 to set a predetermined orientation. In this embodiment, the 3D image of one animal extracted by the individual target animal image acquisition means 35 using a segmentation method is used to measure each part using the planar dimension / area information acquisition means 33 and the side dimension / area information acquisition means 34. The parts to be measured can be shoulder width information, waist width information, hip width information, length from shoulder to hip, and the area of ​​the entire body. To measure each part, it is easier to calculate by slicing the image of the pig into cross-sections at regular intervals and processing them sequentially from the head, as described above. However, in this embodiment, since the pigs are photographed as they come to eat, images with various orientations are obtained. The posture is also not consistent, so ingenuity is required. The processing performed by the correction processing means 38, including solutions to these problems, will be described below.

[0034] <Correction process> If the image of the pig is cleanly cut out, a 3D image can be binarized and an ellipse approximation can be performed along the body's boundaries. The center of this ellipse will correspond to the center of the body, and the angle of the ellipse will directly match the angle of the pig. Therefore, we calculate the angle of the ellipse in the state shown in Figure 13(A), and rotate the image using an affine transformation with the center of the ellipse as the axis of rotation so that the angle of the ellipse that makes the image horizontal becomes horizontal (Figure 13(B)).

[0035] The weight management and retention system 200 with the above configuration processes data according to a program corresponding to the flowchart shown in Figure 14. The operation will be described below based on the flowchart. The CPU 10 captures images of the target animal using the camera 26 (S11). Once the camera 26 has captured images of the target animal, the CPU 10 receives the captured image data (S12) and separates the overlapping images of the target animals into images of a single animal (S21).

[0036] Next, the CPU 10 performs individual identification (image recognition) and assigns identification information to each image of the target animal (S22). Then, it discards images that are distorted and selects only images that are captured in good condition (S23). The CPU 10 also performs elliptic approximation along the body interface of the target animal, calculates the angle of the ellipse, and corrects the image by rotating it using an affine transformation with the center of the ellipse as the axis of rotation so that the angle of the ellipse that makes the image horizontal becomes horizontal (S24). Furthermore, it obtains planar width information, planar length information, and side dimensions / width information (S13). The CPU 10 provides the obtained planar width information, planar length information, and side dimensions / width information to a machine learning model to obtain the weight of the target animal (S14). In this way, it becomes possible to appropriately measure the weight of the target animal.

[0037] Furthermore, the weight information acquisition means 40 may calculate the average value of the weight information obtained within a predetermined time for a target animal identified as a single individual, and use the calculated average value as the weight of that target animal. Also, even if the target animal remains within the field of view, there may be cases where weight estimation is not possible due to poor conditions such as posture, which may result in fewer weight measurements. If the acquired weight measurement value is below a set threshold, the acquired data may be discarded. Alternatively, if weight measurements are obtained more than the threshold number of times, the data may be averaged and the process completed as the weight measurement value for one target animal.

[0038] Furthermore, the weight management and holding system 200 may be equipped with a program that implements a species determination means (not shown) that determines species information within the target animal based on the above planar dimension / area information and the above side dimension / area information. In this case, the individual information management means 352 performs the operation of providing the planar dimension / area information acquisition means 33, the side dimension / area information acquisition means 34, and the weight information acquisition means 40 with an image that has identification information attached along with the species information determined by the species determination means, and the weight information acquisition means 40 may adjust the weight information using parameters corresponding to the species information within the target animal when obtaining weight information.

[0039] Each of the above embodiments reduces the effort required to measure weight, enables daily management and data collection of pig weights, and can lead to maximizing the profits of pig farms by optimizing the timing of shipments, for example. Furthermore, by setting the threshold to 0 kg, it is possible to measure the weight of all pigs in the pen without duplication. Since the pigs' feed is changed according to their weight, by observing the pigs' growth progress (weight), it is possible to determine the timing for changing feed according to their growth (weight). This applies not only to pigs but also to other types of animals. In other words, the task of appropriately setting a threshold for the weight of an animal and marking animals within a desired range or time period can be performed appropriately without requiring manual labor or time. In the weight management and retention system 200 of the second embodiment, the weight of the target animal is measured, and set information consisting of individual identification information, weight information, and image information used for individual identification is stored in the storage 23.

[0040] Depending on the configuration described above in the weight management and retention system 200 of the first embodiment and the weight management and retention system 200 of the second embodiment, the following technical challenges arise. Specifically, are there any problems in actual operation with the accuracy of these embodiments? Also, can the same accuracy be ensured when deploying the system of this embodiment to various farms? Furthermore, the accuracy of the machine learning model 41 was estimated as shown in Figure 15. That is, the mean squared error (RMSE) was 2.6%. From this perspective, can the first embodiment of the present invention be operated with the same accuracy even if the type of actual product changes? In view of these challenges, the weight management and retention system 200 of a certain machine learning model configuration and the weight management and retention system 200 of the second embodiment are equipped with a calibration unit 60. The calibration unit 60 will be described below.

[0041] The calibration unit 60 is equipped with a history information storage means 61, an updated learning model generation means 62, and an update timing control means 63. The history information storage means 61 stores pairs of explanatory variables and target variables as history information when weight information is acquired using the machine learning model 41 in the weight information acquisition means 40. The pairs of explanatory variables and target variables are stored in association with identification information issued for each individual by the individual information management means 32.

[0042] The updated learning model generation means 62 obtains an updated machine learning model by performing the same machine learning as when the machine learning model 41 was created, based on the input of the actual weight information of the target animal corresponding to the explanatory variables in the pairs of historical information stored in the historical information storage means 61, and using this actual weight information as an explanatory variable and the actual weight information as an objective variable. More specifically, the updated machine learning model may be obtained when a predetermined number of the actual weight information to be used as the objective variable is available. In other words, for example, the weight of the target animal can be measured using a scale when it is shipped, and the data can be input to the calibration unit 60 in association with the identification information to create pairs of objective variables and explanatory variables to be used in generating the updated machine learning model. Alternatively, it is also possible to use identification information from a meat processing plant and weight information obtained by measuring the carcass, which is the meat obtained by processing one individual animal.

[0043] Furthermore, if a sufficient number of target-and-explanatory variable pairs for generating the updated machine learning model cannot be obtained, the target-and-explanatory variable pairs created from the actual weight information may be mixed with the target-and-explanatory variable pairs used to generate the machine learning model 41. Also, if the weight information acquisition means 40 is used to adjust parameters according to the species information within the target animals when obtaining weight information, calibration may be performed to change these parameters according to the actual weight information. Moreover, calibration may be performed on the machine learning model 41 of the farm being calibrated using target-and-explanatory variable pairs created from actual weight information obtained from another farm with similar rearing methods and pig species ratios.

[0044] An update timing control means 63 is provided to determine the error between the weight information obtained using the machine learning model and the actual weight information. The update timing control means 63 creates error information between the weight information obtained using the machine learning model 41 and the actual weight information, and controls the timing of acquiring the updated machine learning model based on this error information. Therefore, the update timing control means 63 can be used to create and output error information between the weight information obtained using the machine learning model 41 and the actual weight information, and to monitor the error.

[0045] The update timing control means 63 can signal that it is time for an update when the mean squared error (RMSE) shown in Figure 15 exceeds, for example, 3%. Alternatively, the update learning model generation means 62 may be made to operate using pairs of target and explanatory variables created from the collected physical weight information, and the generated update learning model may be used as the new machine learning model 41.

[0046] According to the calibration unit 60, it is expected that the accuracy of this embodiment will be sufficient for actual operation, that the same accuracy can be ensured when deploying the system of this embodiment to various farms, and that the same accuracy can be maintained even if the breed of pig changes. The calibrated weight information is also stored in the storage 23 as set information, which includes individual identification information and image information used for individual identification.

[0047] Next, the mark application system 100 will be described. Figure 16 shows a system configuration diagram centered on the edge computer 150 that constitutes the mark application system 100. In other words, the CPU 310 configures the mark application system 100 using programs and data in the main memory 311. The CPU 310 is connected to a storage interface 313, an input interface 314, an output interface 315, and a data input interface 316 via a bus 312.

[0048] A storage device 323 is connected to the storage interface 313. The storage device 323 stores programs and data necessary for the operation of the mark application system 100, which the CPU 310 can read and use from the main memory 311 as needed. The storage device 323 includes external storage devices such as HDDs, auxiliary storage devices, and cloud storage. The storage device 323 stores programs that implement the means described later. An output interface (not shown) of the weight management and holding system 200 is connected to the input interface 314, and the system is configured to allow the mark application system 100 to receive set information, which is a set of identification information including individual identification information, weight information, and image information used for individual identification, stored in the storage device 23, via this output interface.

[0049] The mark attachment mechanism 170 is connected to the output interface 315. Cameras 326-1 to 326-m, which are mark attachment area cameras 160, are connected to the data input interface 316, and these cameras 326-1 to 326-m function to image the target animal. The number of cameras 326-1 to 326-m is arbitrary and can be configured as a 3D camera. Image data obtained by cameras 326-1 to 326-m is taken up by the data input interface 316 to the CPU 310, storage 323, and other required parts.

[0050] In an embodiment of the mark application system 100, as shown in Figure 17, the system includes a camera 326 (326-1 to 326-m), an acquisition means 211, a determination means 212, a control means 213, a predetermined color / predetermined area mark detection means 214, and a distance / application position detection means 215. These means can be implemented by the CPU 310 appropriately reading programs from the storage 323 into the main memory 311, or by using programs that are initially stored in the main memory 311.

[0051] The mark attachment area is created by the feeder 410 shown in Figure 18. Specifically, the feeder 410 has a hopper 411 that is the shape of an inverted truncated cone for accumulating food, and a cylindrical feeding area 412 is provided at the bottom of the hopper 411. The feeding area 412 has a donut-shaped planar shape as shown in Figure 19. The feeding area 412 is divided into multiple (in this case, 8) individual feeding areas 415 by the fence bars 414, which are stretched from a circular frame 413 surrounding the lower part of the hopper 411 toward the outer frame of the feeding area 412.

[0052] Figures 20 and 21 show another example of the feeder 410. The feeder 410 has a hopper 411 with an inverted wedge shape for accumulating food, and a rectangular feeding area 412 with an elongated planar shape is provided at the bottom of the hopper 411. One side of the hopper 411 is an inclined surface 421, and a wall 423 extending from this inclined surface 421 toward the bottom surface 422 divides the feeding area 412 into multiple individual areas 415 where one individual eats food.

[0053] In the feeder 410 shown in Figures 20 and 21, the edge computer 150 is installed on the top plate of the hopper 411. The coating liquid tank 171, which is part of the marking mechanism 170, is located on the top plate of the hopper 411. A water gun-type marking attachment (not shown) is connected to the coating liquid tank 171, and an outlet 172 for spraying the coating liquid is installed at a position on the upper side wall of the hopper 411 overlooking the individual area 415. The optimal position for the outlet 172 is 1.2m or higher, which is a height that prevents the pigs from standing up and causing mischief, but it can be set to a height of 1.0m to 1.2m considering the height of the animals. Here, we used a water gun-style marking attachment, but this is not the only option. In other words, spray cans, paint spray guns, or air guns can also be used. However, spray cans have the disadvantage of needing to be shaken, and depending on the size of the can, frequent replacement may be necessary. Paint spray guns require spraying at close range. Furthermore, air guns require a compressor, resulting in a larger and more elaborate facility.

[0054] In the feeder 410 shown in Figures 18 and 19, as shown in Figure 22, the coating liquid tank 171 included in the edge computer 150 and mark application mechanism 170 is provided on the top plate of the hopper 411. An injector 172 for injecting the coating liquid is installed at a position on the upper side wall of the hopper 411 overlooking the individual area 415.

[0055] The marking system 100, which includes a marking mechanism 170 having the above configuration, functions as follows, as shown in Figure 17. The acquisition means 211 uses the image information of the animal obtained by the marking area camera 160 and the set information managed and held by the weight management and holding system 200 to acquire individual identification information and weight information of the animal related to the image information currently obtained by the marking area camera 160. The determination means 212 has weight threshold information for whether or not to apply a mark, and uses the weight information obtained by the acquisition means 211 to determine whether or not to apply a mark. The control means 213 controls the marking mechanism 170 based on the determination result by the determination means 212 to perform or not perform the application of a mark to the animal in the marking area.

[0056] The predetermined color / predetermined area mark detection means 214 sets a binding box on the image of the animal obtained by the mark attachment area camera 160 to obtain image information of a predetermined range, and detects a mark of a predetermined color and area within this predetermined range of image information. The area is determined by an HSV mask. In response to the operation of the predetermined color / predetermined area mark detection means 214, the control means 213 operates as follows: That is, the control means 213 controls whether or not to attach a mark to the animal based on the result of the predetermined color / predetermined area mark detection means 214.

[0057] For example, if the image of an animal obtained by the mark application area camera 160 is as shown in Figure 23(a), and three red marks R are applied to the binding box B, an image like Figure 23(b) is obtained, the red color is detected and the area of ​​the three points is calculated, and if it is a predetermined color and the area is greater than or equal to a predetermined value, the system controls whether or not to apply the marks. Otherwise, the system controls whether or not to apply the marks.

[0058] Furthermore, the distance / adhesion position detection means 215 obtains 3D image information based on the image information of the animal and the mark adhesion mechanism 170 obtained by the mark adhesion area camera 160, and determines the distance between the nozzle 172 and the animal and the adhesion position on the animal to which the coating liquid adheres. In response to the operation of the distance / adhesion position detection means 215, the control means 213 operates as follows: That is, the control means 213 controls whether or not to apply marks to the animal based on the results of the distance / adhesion position detection means 215.

[0059] The mark attachment system 100 with the above configuration processes data according to a program corresponding to the flowchart shown in Figure 24. The operation will be described below based on the flowchart. The CPU 310 captures an image of the target animal using the camera 326 (S51). It then acquires the captured image information (S52). Next, using the acquired image information and the set information managed and held by the weight management and holding system 200, it acquires the individual identification information and weight information of the animal related to the image information currently obtained by the mark attachment area camera 160 (S53).

[0060] For example, when there is currently obtained image information as shown on the left of Figure 25, and set information is provided as shown in the table on the right of Figure 25, image information matching is performed to obtain the individual identification information and weight of the set information that has matching image information. In the example of Figure 25, the individual identification information is obtained as 0003 and the weight as 120 kg.

[0061] Next, based on the weight threshold information for determining whether or not a mark should be applied, the weight information obtained above is used to determine whether or not a mark should be applied. The determination of whether or not a mark should be applied is made based on whether a mark of a predetermined area and a predetermined color could be detected in the image information within a predetermined range (S54). If the result is NO here, the process returns to step S51 and continues.

[0062] If the answer in step S54 is YES, then it is determined whether a mark should be applied based on whether a mark of a predetermined color and area was detected in the image information within a predetermined range (S55). If the answer here is NO, the process returns to step S51 and continues.

[0063] If the answer in step S55 is YES, 3D image information is obtained based on the image information to determine the distance between the nozzle and the animal and the attachment position on the animal where the coating liquid adheres, and it is detected whether the determined position is appropriate (S56). If the answer here is NO, the process returns to step S51 and continues.

[0064] If the answer in step S56 is YES, the mark attachment mechanism is controlled to attach the marks to the animals in the mark attachment area (S57), and the process returns to step S51 to continue.

[0065] In this way, it is possible to perform the marking of target animals that have reached the desired weight appropriately, without requiring manpower or time. In this embodiment, the marking system 100 and the weight management and holding system 200 are provided as separate computers, but they may be combined into one. [Explanation of Symbols]

[0066] 10 CPU 11 Main memory 12 buses 13 Storage Interfaces 14 Input Interfaces 15 Display Interface 16. Data Input Interface 22 Pointing devices 23 Storage 24 Input devices 25 Display device 26 cameras 31 Image acquisition method 32 Individual information management means 33. Means for acquiring planar dimensions and area information 34. Means for acquiring side dimensions and width information 35. Means for acquiring images of individual target animals 37. Appropriate Image Selection Method 38 Correction processing means 40. Means for acquiring weight information 41 Machine Learning Models 60 Calibration section 61. Means for storing historical information 62 Update learning model generation means 63 Update timing control means 100 Mark Application System 150 Edge Computers 160 Marked Area Camera 170 Marking Mechanism 171 Coating liquid tank 172 Injection port 200 Weight Management and Maintenance System 211 Acquisition means 212 Judgment means 213 Control means 214 Predetermined color / predetermined area mark detection means 215 Distance and attachment position detection means 250 Edge Computers 260 Weight Measurement Camera 310 CPU 311 Main Memory 312 Bus 313 Storage Interfaces 314 Input Interfaces 315 Output Interface 316 Data Input Interface 323 Storage 326 Camera 352 Individual information management means 365 Individual Target Animal Image Acquisition Method 410 Feeder 411 Hoppa 412 Feeding Area 413 yen frame 414 Fence pole

Claims

1. This system obtains image information of animals from a weight-measuring camera, performs individual animal identification based on this image information, measures the weight of the animal, and works in cooperation with a weight management and retention system that manages and maintains set information consisting of individual identification information, weight information, and the image information used for individual identification. A mark-attachment area camera for obtaining image information of the animals located in the mark-attachment area, A marking mechanism for attaching marks to animals in the aforementioned marking area, A marking system comprising the above, which applies marks to animals in the marking area, An acquisition means that uses the image information of the animal obtained by the mark-attached area camera and the set information managed and maintained in the weight management and retention system to acquire individual identification information and weight information of the animal related to the image information currently obtained by the mark-attached area camera, A determination means that has weight threshold information for determining whether or not a mark should be attached, and determines whether or not a mark should be attached using the weight information obtained by the acquisition means, A control means controls the mark attachment mechanism based on the determination result of the determination means to perform or not perform the application of marks to animals in the mark attachment area. A mark application system characterized by comprising the following:

2. The marking system according to claim 1, wherein the acquisition means searches for image information of an animal that matches the animal obtained by the marking area camera by image recognition using the image information of the animal obtained by the marking area camera and the image information in the set information managed and held in the weight management and holding system, and acquires the individual identification information and weight information of the animal.

3. The system includes a mark detection means that sets a binding box on an image of an animal obtained by the mark attachment area camera to obtain image information of a predetermined range, and detects a mark of a predetermined color and area in this predetermined range of image information, The mark application system according to claim 1, characterized in that the control means controls whether or not to apply a mark to an animal based on the result of the predetermined color and predetermined area mark detection means.

4. The mark application mechanism has an injection port for ejecting the coating liquid, The system includes a distance / adhesion position detection means that obtains 3D image information based on the image information of the animal and the mark adhesion mechanism obtained by the mark adhesion area camera, and determines the distance between the nozzle and the animal and the adhesion position on the animal to which the coating liquid adheres. The mark attachment system according to claim 1, characterized in that the control means controls whether or not to attach a mark to an animal based on the results of the distance and attachment position detection means.

5. The marked area is the feeding area where one animal eats food in the feeder. The marking system according to claim 1, characterized in that the marking area camera is attached to the feeder and configured to photograph animals in each feeding area from above.

6. The mark attachment system according to claim 5, characterized in that the mark attachment area is configured such that the feeding area, which has a donut-shaped planar shape centered on the feeder, is divided into individual areas where multiple individuals eat food.

7. The mark attachment system according to claim 5, characterized in that the mark attachment area is configured such that the rectangular feeding area in plan view is arranged in a manner in which multiple individual areas where one individual eats food are lined up.

8. This system works in conjunction with a weight management and retention system that obtains image information of animals from a weight measurement camera, performs individual animal identification based on this image information, measures the weight of the animal, and manages and maintains set information consisting of individual identification information, weight information, and the image information used for individual identification. A mark-attaching area camera for obtaining image information of animals in the marked area, A marking mechanism for attaching marks to animals in the aforementioned marking area, A computer is provided in a mark application system that applies marks to animals in the mark application area, An acquisition means that uses the image information of the animal obtained by the mark-attached area camera and the set information managed and maintained in the weight management and retention system to acquire individual identification information and weight information of the animal related to the image information currently obtained by the mark-attached area camera. A determination means that has weight threshold information for determining whether or not a mark should be attached, and determines whether or not a mark should be attached using the weight information obtained by the acquisition means. A control means that controls the mark attachment mechanism based on the determination result of the determination means to perform or not perform the application of marks to animals in the mark attachment area. A mark application program characterized by its function as such.

9. The mark-attaching program according to claim 8, characterized in that the computer is used as the acquisition means to perform image recognition using the image information of the animal obtained by the mark-attaching area camera and the image information in the set information managed and maintained in the weight management and holding system to search for image information of an animal that matches the animal obtained by the mark-attaching area camera, and to acquire the individual identification information and weight information of the said animal.

10. The aforementioned computer A binding box is set on the image of the animal obtained by the mark-attaching area camera to obtain image information of a predetermined range, and this is used as a predetermined color / predetermined area mark detection means to detect a mark of a predetermined color and area on the image information of this predetermined range. The mark application program according to claim 8, characterized in that the computer is used as the control means to control whether or not to apply a mark to an animal based on the results of the predetermined color and predetermined area mark detection means.

11. The mark application mechanism has an injection port for ejecting the coating liquid, The aforementioned computer Based on the image information of the animal and the mark attachment mechanism obtained by the mark attachment area camera, 3D image information is obtained and this is used as a distance / attachment position detection means to determine the distance from the nozzle to the animal and the attachment position on the animal to which the coating liquid adheres. The marking program according to claim 8, characterized in that the computer is used as the control means to control whether or not to perform marking on an animal based on the results of the distance and attachment position detection means.

12. This system works in conjunction with a weight management and retention system that obtains image information of animals from a weight measurement camera, performs individual animal identification based on this image information, measures the weight of the animal, and manages and maintains set information consisting of individual identification information, weight information, and the image information used for individual identification. A mark-attaching area camera for obtaining image information of animals in the marked area, A marking mechanism for attaching marks to animals in the aforementioned marking area, In a mark application system that applies marks to animals in the mark application area, the CPU used in the system appropriately reads programs and data for operation from storage into main memory and uses them to process the data, and the mark application method is executed by processing the data, An acquisition step to acquire individual identification information and weight information of an animal related to the image information currently obtained by the mark-attached area camera, using the image information of the animal obtained by the mark-attached area camera and the set information managed and maintained in the weight management and retention system, A determination step that has weight threshold information for determining whether or not a mark should be attached, and uses the weight information obtained in the acquisition step to determine whether or not a mark should be attached, A control step which controls the mark attachment mechanism based on the determination result from the determination step to perform or not perform the act of attaching marks to animals in the mark attachment area. A method for attaching a mark, characterized by comprising the following:

13. The marking method according to claim 12, characterized in that in the acquisition step, image recognition is performed using the image information of the animal obtained by the marking area camera and the image information in the set information managed and maintained in the weight management and holding system to search for image information of an animal that matches the animal obtained by the marking area camera, and to acquire the individual identification information and weight information of the said animal.

14. The system includes a predetermined color and predetermined area mark detection step in which a binding box is set on the image of the animal obtained by the mark attachment area camera to obtain image information of a predetermined range, and a predetermined color and predetermined area mark is detected on this predetermined range of image information, The mark application method according to claim 12, characterized in that the control step controls whether or not to apply the mark to the animal based on the detection result in the predetermined color and predetermined area mark detection step.

15. The mark application mechanism has an injection port for ejecting the coating liquid, The system includes a distance and attachment position detection step that obtains 3D image information based on the image information of the animal and the mark attachment mechanism obtained by the mark attachment area camera, and determines the distance between the nozzle and the animal and the attachment position on the animal to which the coating liquid adheres. The marking method according to claim 12, characterized in that the control step controls whether or not to perform marking on an animal based on the detection result of the distance and attachment position detection step.

Images