Weight Estimation Device

The body weight estimation device uses a one-shot 3D scanner and regression equations to overcome inaccuracies in conventional methods, achieving high-accuracy weight estimation by averaging three-dimensional information and selecting appropriate correlations.

JP7579010B2Active Publication Date: 2024-11-07NOA CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2023036207
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2018-04-25
Filing Date
2023-03-09
Publication Date
2024-11-07
Estimated Expiration
2039-03-25

AI Technical Summary

Technical Problem

Conventional methods for measuring livestock weight, such as using a scale or optical video cameras, result in inaccurate weight estimation due to two-dimensional information limitations, leading to large measurement errors.

Method used

A body weight estimation device utilizing a one-shot 3D scanner to acquire three-dimensional information, a calculation device to determine geometric quantities, and a storage unit with regression equations to estimate weight based on the animal's external shape, allowing for high-accuracy weight estimation.

Benefits of technology

The device provides accurate weight estimation with a simple configuration by averaging multiple shots and selecting appropriate regression equations based on animal characteristics.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007579010000001
    Figure 0007579010000001
  • Figure 0007579010000002
    Figure 0007579010000002
  • Figure 0007579010000003
    Figure 0007579010000003
Patent Text Reader

Abstract

To provide a weight estimation device that can estimate the weight of an animal with high accuracy despite having a simple configuration. [Solution] The weight estimation device is a weight estimation device that estimates the weight of an animal, and includes an information acquisition device which is a one-shot 3D scanner that images the animal and acquires three-dimensional information for each of multiple points, a calculation device that calculates the geometric quantities of the animal's external shape based on average three-dimensional information obtained by averaging multiple pieces of three-dimensional information acquired by the one-shot 3D scanner by taking multiple shots within a predetermined time, and a weighing device that estimates the weight of the imaged animal based on the geometric quantities of the animal's external shape calculated by the calculation device.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a body weight estimation device capable of estimating the body weight of an animal. [Background technology]

[0002] By measuring the weight of livestock, the state of their growth can be understood and the appropriate time for shipping can be predicted. Conventionally, livestock have been weighed by placing them on a scale. However, it is time-consuming to place each livestock on the scale one by one. In addition, there is also the problem that the livestock move on the scale, making it difficult to measure their weight accurately.

[0003] In response to this, Patent Document 1 discloses a technology in which an image of an animal is taken using an optical imaging camera fixed above, and the animal's body height and projected area are determined based on the image, thereby calculating the animal's weight. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2002-286421 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, the technology of Patent Document 1 has a problem in that only two-dimensional information such as the body height and projected area of ​​an animal can be obtained, resulting in large measurement errors and making it difficult to measure the weight of an animal with high accuracy.

[0006] The present invention has been made in consideration of the problems with the conventional technology, and has an object to provide a body weight estimation device that is simple in configuration but can estimate the body weight of an animal with high accuracy. [Means for solving the problem]

[0007] The body weight estimation device of the present invention is a body weight estimation device that estimates the body weight of an animal, An information acquisition device that is a one-shot type 3D scanner that captures an image of the animal and acquires three-dimensional information for each of a plurality of points; A calculation device that calculates a geometric quantity of the outer shape of the animal based on average three-dimensional information obtained by averaging a plurality of pieces of three-dimensional information acquired by the one-shot type 3D scanner performing a plurality of shots within a predetermined time; a weighing device that estimates the weight of the imaged animal based on the geometric quantities of the animal's external shape calculated by the calculation device; 、 A memory unit that stores a regression equation that represents a correlation between the geometric quantity of the outer shape of the animal and the weight of the animal, the weighing device estimates the weight of the animal by substituting the geometric quantity of the external shape of the animal calculated by the calculation device into the regression equation; The memory unit stores a plurality of regression equations that represent a plurality of different correlations according to characteristics of an animal to be estimated, The weighing device selects one of the regression equations corresponding to the geometric quantities of the external shape of the animal calculated by the arithmetic device, reads it from the storage unit, and estimates the weight of the animal by substituting the geometric quantities of the external shape of the animal calculated by the arithmetic device. It is characterized by: The body weight estimation device of the present invention is a body weight estimation device that estimates the body weight of an animal, An information acquisition device that is a one-shot type 3D scanner that captures an image of the animal and acquires three-dimensional information for each of a plurality of points; A calculation device that calculates geometric quantities of the outer shapes of the animals based on a plurality of pieces of three-dimensional information acquired by the one-shot 3D scanner performing a plurality of shots within a predetermined time; a weighing device that estimates the weight of the imaged animal based on an average geometric quantity obtained by averaging the geometric quantities of the external shapes of a plurality of the animals; 、 A memory unit that stores a regression equation that represents a correlation between the geometric quantity of the outer shape of the animal and the weight of the animal, the weighing device estimates the weight of the animal by substituting the geometric quantity of the external shape of the animal calculated by the calculation device into the regression equation; The memory unit stores a plurality of regression equations that represent a plurality of different correlations according to characteristics of an animal to be estimated, The weighing device selects one of the regression equations corresponding to the geometric quantities of the external shape of the animal calculated by the arithmetic device, reads it from the storage unit, and estimates the weight of the animal by substituting the geometric quantities of the external shape of the animal calculated by the arithmetic device. It is characterized by: The body weight estimation device of the present invention is a body weight estimation device that estimates the body weight of an animal, An information acquisition device that is a one-shot type 3D scanner that captures an image of the animal and acquires three-dimensional information for each of a plurality of points; A calculation device that calculates geometric quantities of the outer shapes of the animals based on a plurality of pieces of three-dimensional information acquired by the one-shot 3D scanner performing a plurality of shots within a predetermined time; a weighing device that calculates a plurality of weights of the imaged animals based on the geometric quantities of the external shapes of the animals, and averages these weights to obtain an estimated weight; 、 A memory unit that stores a regression equation that represents a correlation between the geometric quantity of the outer shape of the animal and the weight of the animal, the weighing device estimates the weight of the animal by substituting the geometric quantity of the external shape of the animal calculated by the calculation device into the regression equation; The memory unit stores a plurality of regression equations that represent a plurality of different correlations according to characteristics of an animal to be estimated, The weighing device selects one of the regression equations corresponding to the geometric quantities of the external shape of the animal calculated by the arithmetic device, reads it from the storage unit, and estimates the weight of the animal by substituting the geometric quantities of the external shape of the animal calculated by the arithmetic device. It is characterized by: Effect of the Invention

[0008] According to the present invention, it is possible to provide a body weight estimation device that can estimate the body weight of an animal with high accuracy while having a simple configuration. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a block diagram of a body weight estimation device according to the present embodiment. [Diagram 2] FIG. 2 is a diagram showing the body weight estimation device 1 in use. [Diagram 3] FIG. 3 is a diagram showing a state in which the body weight estimation device 1 according to the modified example is used. [Figure 4] FIG. 4 is a schematic diagram showing a subject based on three-dimensional information obtained by the imaging unit 5. As shown in FIG. [Diagram 5] FIG. 5 is a schematic diagram showing a state in which animals whose body weights have been estimated are extracted. [Figure 6] Figure 6 is a graph showing the correlation between body weight on the vertical axis and lower body surface area on the horizontal axis. [Figure 7] FIG. 7 is a diagram showing data in which the surface area A of the lower body of a pig is plotted in correspondence with its actual weight W. [Figure 8] FIG. 8 is a diagram showing data in which the body lengths L of pigs raised in the first pigpen are plotted in association with their actual weights W. [Figure 9] FIG. 9 is a diagram showing data in which the body lengths L of pigs raised in the second pigpen are plotted in association with their actual weights W. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] The present embodiment will be described with reference to the drawings. Fig. 1 is a block diagram of a body weight estimation device according to the present embodiment. Fig. 2 is a diagram showing a state in which the body weight estimation device 1 is in use. In this embodiment, the term "imaging" refers to at least obtaining three-dimensional coordinate information of an imaging target.

[0011] In FIG. 1, a body weight estimation device 1 includes an operation unit 3 provided in a housing 2, a display unit 4, an imaging unit (information acquisition device) 5, a storage unit 6, and a CPU (arithmetic unit and weighing device) 7 that controls these components.

[0012] The operation unit 3 is an interface that the user operates to input instructions to the body weight estimation device 1. In FIG. 2, a power switch 3a and a release button 3b are illustrated as the operation unit 3, but a touch panel may be provided instead of the release button. In this case, a touch sensor is provided on the surface of the display unit 4, and the user can input instruction information to the body weight estimation device 1 by touching the screen. In such a case, the display unit also serves as the operation unit.

[0013] The display unit 4 has a function of displaying images and information, which can be visually recognized by the user. A liquid crystal display can be used as the display unit 4.

[0014] The CPU 7 reads and executes the programs stored in the storage unit 6, and performs operations on various data, generates control signals, and so on.

[0015] The storage unit 6 provides a function of storing various data used in the body weight estimation device 1. Examples of the storage unit 6 include a RAM or buffer used as a temporary working area for the CPU 7, a read-only ROM, a non-volatile memory (such as a NAND memory), a portable storage medium (such as an SD card or a USB memory) attached to a dedicated reading device, and the like.

[0016] The imaging unit 5 is a camera that acquires three-dimensional information, and captures an animal as a subject to acquire three-dimensional subject information having three-dimensional coordinate information for each pixel and color information. Based on the color information, the display unit 4 can display a color image of the subject. Three-dimensional subject information may be used instead of a color image. A known camera can be used to acquire three-dimensional information. The device for acquiring three-dimensional information may be one that uses either a time of flight method or a light pattern irradiation method. However, it is preferable that the shooting time for acquiring one data is short. As a camera for acquiring three-dimensional information, it is most preferable to use a one-shot type 3D scanner. In addition, the imaging unit 5 outputs three-dimensional information obtained by one shot, but it is also possible to perform multiple shots within a predetermined time and output multiple pieces of three-dimensional information obtained by the multiple shots.

[0017] In estimating an animal's body weight, it is desirable to average the acquired data by using one of the following methods. (1) The CPU 7 as a calculation device calculates the geometric quantities of the external shape of the animal based on average three-dimensional information obtained by averaging a plurality of pieces of three-dimensional information, and determines the estimated body weight. (2) The CPU 7 as a calculation device calculates multiple geometric quantities for the external shape of the same animal based on multiple pieces of three-dimensional information, and estimates the weight of the imaged animal based on an average geometric quantity obtained by averaging the geometric quantities. (3) The CPU 7 as a calculation device calculates multiple geometric quantities for the external shape of the same animal based on multiple pieces of three-dimensional information, calculates multiple weight values ​​for the imaged animal based on the geometric quantities, and averages these weight values ​​to determine the estimated weight.

[0018] Next, the operation of the body weight estimation device 1 will be described. As shown in Fig. 2, when a user holds the housing 2 of the body weight estimation device 1 in his / her hand HD and turns on the power switch 3a, the CPU 7 displays a cross mark MK in the center of the display unit 4 and displays an image of the subject captured by the imaging unit 5 on the display unit 4. The display position of the cross mark MK on the screen is not limited to the center. Also, multiple cross marks may be displayed. The CPU 7 and the display unit 4 form a selection device.

[0019] While moving their hand HD, the user moves the housing 2 so that a cross mark is positioned on the flank of the animal AN whose weight is to be estimated on the display unit 4, and the animal AN is imaged in a specified state by the imaging unit 5. At this time, there is no particular restriction on the distance from the animal AN to the user, but the entire animal AN should be captured.

[0020] When the user operates the release button 3b with the cross mark MK positioned on the flank of the animal AN, the imaging unit 5 obtains three-dimensional subject information and color information including the animal AN, and stores them in the storage unit 6. Furthermore, the CPU 7 estimates the weight of the animal AN based on the three-dimensional subject information.

[0021] 3 is a diagram showing a state in which a body weight estimation device 1 according to a modified example is in use. In this modified example, operation units such as a power switch 3a and a release button 3b are disposed on a grip section 2a connected to the bottom surface of the housing section 2. This allows a user to hold the body weight estimation device 1 simply by grasping the grip section 2a with one hand, and to operate the body weight estimation device 1 using only that one hand.

[0022] The reason why it is preferable to capture an image while holding the body weight estimation device 1 in only one hand will be explained. Even if the body weight of an animal is estimated by operating the body weight estimation device 1 as described above, if the target animal moves, it may become difficult to distinguish it from other animals and it may be lost. Therefore, it is desirable to perform an operation such as painting the target animal at the same time as capturing the image.

[0023] According to this modified example, weight estimation device 1 can be held in only one hand while capturing an image, leaving the other hand free to do things like paint the animal and avoid losing sight of the target animal.

[0024] Furthermore, although not shown, the weight estimation device 1 can be provided with a selection button (selection unit) as an operation unit. Such a selection button can be provided on the screen by making the screen a touch panel. Depending on the type of animal, for example, if the animal to be estimated is a pig that is B months old or older, the first selection button is turned on. Also, if the animal to be estimated is a pig that is A months old or older but less than B months old, the second selection button is turned on. Furthermore, if the animal to be estimated is a pig that is less than A months old, the third selection button is turned on. Depending on this selection, the CPU 7 performs different calculations. The details of the calculations will be described later. The method of dividing the ages is one example.

[0025] The processing performed by the CPU 7 will be described below, taking a pig as an example of an animal for which weight estimation is to be performed. The three-dimensional subject information stored in the storage unit 6 includes, in addition to the animal AN, information on the floor FL and other objects OBJ (including other animals) that are not required for weight estimation processing. Therefore, the CPU 7 extracts only the three-dimensional subject information of the pig for which weight estimation is desired.

[0026] The three-dimensional subject information has three-dimensional coordinates (x, y, z) based on the weight estimation device 1 for each of the countless points (e.g., pixels) on the screen. Here, the x-axis is defined as the horizontal direction on the screen, with the left side being positive, the y-axis is defined as the vertical direction on the screen, with the top being positive, and the z-axis is defined as the perpendicular direction on the screen, with the far side being positive. The CPU 7 then compares the three-dimensional coordinates for each adjacent point, and separates unnecessary three-dimensional subject information from the inherent characteristics of the background.

[0027] More specifically, the CPU 7 specifies the plane that is most frequently on a certain plane among the three-dimensional point cloud acquired as the three-dimensional object information, determines that plane to be the floor FL, and removes these points from the target. Here, the state of being most frequently on a plane means the state in which the number of points within a predetermined distance from the plane is the largest.

[0028] Furthermore, based on the premise that the point corresponding to or adjacent to the cross mark MK is the pig's flank, the CPU 7 compares the three-dimensional coordinates of each adjacent point, recognizes the object that forms a continuous surface as the pig AN, and recognizes the other points as unnecessary objects OBJ, and removes these points from the target. As a result, only the three-dimensional coordinate group corresponding to the target pig AN remains (see FIG. 5).

[0029] Next, the CPU 7 obtains geometric quantities related to the outline of the pig AN from the extracted three-dimensional coordinate group. First, referring to FIG. 5, the total length L of the pig AN is obtained based on the extracted three-dimensional coordinate group. The total length L is defined as the maximum x value minus the minimum value among the x coordinates of the points constituting the pig AN. Next, the pig AN is divided into two at the midpoint P of the total length L, and the surface area is obtained for the left and right sides. The surface area is obtained by forming a triangle by connecting three points in the extracted three-dimensional coordinate group, meshing, and adding up the areas of the triangles. At this time, the influence of noise, etc. can be suppressed by averaging the area according to multiple shots, for example. The meshing is described, for example, in JP 2013-096745 A.

[0030] In the case of the pig AN, the surface area of ​​the lower body is generally larger than that of the upper body. Therefore, the CPU 7 compares the surface area of ​​the left side with that of the right side, and identifies the larger one (the left side in FIG. 5) as the lower body. In this way, the orientation of the pig AN can be identified.

[0031] Another method is to use the integral value of the curvature. By defining the sphere that is most closely located using a group of points that exist within a range a specified distance away from a certain point, and defining the radius of that sphere as the curvature of the target point, a curvature is assigned to each of the point cloud data of the pig AN. As described above, the pig AN is divided into two at the midpoint P of the total length L of the pig AN, and the integral value of the curvature is calculated for each. In general, in the case of the pig AN, the integral value of the curvature of the upper body, which includes the face, is greater than the integral value of the curvature of the lower body, so the part with the smaller integral value of the curvature (the left side in Figure 5) is identified as the lower body.

[0032] The accuracy of meshing the surface of the pig AN is greatly affected by the imaging distance between the pig AN and the weight estimation device 1. Therefore, it is desirable to perform a smoothing process before meshing. Specifically, when meshing, if the imaging distance is close, the meshing parameters (search distances for points constituting a triangle) are made small, whereas if the imaging distance is far, a correction is made to increase the meshing parameters, thereby enabling appropriate meshing.

[0033] As a result of the inventor's investigation, it has been found that the weight can be estimated using any of the geometric quantities, namely the total length L of the pig AN, the height of the pig's body at the midpoint P, the maximum circumference T of the body at the midpoint P, and the surface area A. However, it has also been found that the correlation between the surface area of ​​the body excluding the head of the pig AN and the weight is the most likely. Therefore, here, the surface area of ​​the lower body of the pig AN will be used as a parameter to estimate the weight. Note that the surface area does not necessarily have to be defined as the surface area of ​​the lower body from the midpoint P, and it is sufficient that the head part is not included.

[0034] Prior to this estimation, the memory unit 6 determines the correlation between the surface area of ​​the lower body of a typical pig and its weight. For example, as shown in Fig. 6, when the measured weights W of a given number of samples of pigs are plotted on the vertical axis and the corresponding surface areas A of the lower body on the horizontal axis, this correlation can be approximated by a regression line (regression equation) W = aA + b. The coefficients a and b can be determined from the results of a survey or simulation carried out in advance, and are stored in the memory unit 6.

[0035] Therefore, when the CPU 7 calculates the surface area A of the lower body of the pig AN, the weight W can be calculated with high accuracy from the regression line W=aA+b. The calculated weight W can be displayed on the display unit 4 as "123 kg" together with an image of the pig AN, for example, as shown in Fig. 2. As mentioned above, it is desirable to average the obtained data when estimating the weight.

[0036] On the other hand, more accurate weight estimation can be achieved by using other geometric quantities as parameters in addition to the surface area of ​​the lower body of the pig AN. For example, if the maximum trunk circumference T is used as an additional parameter, the weight of the pig AN can be estimated with higher accuracy.

[0037] In this case, if the correlation between the surface area A of the pig's lower body, the maximum circumference T of the trunk, and its weight W is found, this correlation can be expressed by a regression line W=aA+bT+c. In this case, the memory unit 6 stores the coefficients a, b, and c.

[0038] In this example, when the CPU 7 calculates the surface area A of the lower body and the maximum trunk circumference T of the pig AN, the weight W can be obtained with high accuracy from the regression line W=aA+bT+c.

[0039] Furthermore, according to the results of the study by the present inventors, it is also found that it is desirable to change the coefficient of the regression line according to the characteristics of the animal. For example, in the case of a piglet, the ratio of the face surface area to the lower body surface area is large, but as the pig grows, the body shape changes so that the ratio of the face surface area to the lower body surface area becomes smaller. Therefore, if the regression line used to estimate the weight is changed according to the pig's age in months, the weight can be estimated more accurately.

[0040] Figure 7 shows data plotting the surface area A of the lower body of a pig in correspondence with its measured weight W, with the regression line of the data also superimposed. Here, when dividing the pigs into data group 3E for ages less than A months, data group 3D for ages A months or more but less than B months, and data group 3C for ages B months or more, it was found that the regression lines differ for each group, as shown below. Note that the unit of weight W is kg, and the unit of surface area A is cm. 2 It is. ·Regression line for pigs less than A months old: W=0.5·A-47 (1) Regression line for pigs aged A months or older and B months or younger: W=0.4·A-21 (2) Regression line for pigs aged B months or older: W=0.3 A+10 (3)

[0041] As described above, the weight estimation device 1 can be provided with a selection button for age in months. Since pigs are categorized and raised according to their age in months, the user can select one of the selection buttons for each breeding location. When one of the selection buttons is selected, the weight estimation device 1 captures an image of the pig, and the CPU 7 selects one of the regression lines of formulas (1) to (3) corresponding to the selected selection button. Furthermore, the CPU 7 can estimate the weight with high accuracy by substituting the obtained surface area of ​​the lower body into the selected regression line. Alternatively, since the image is captured by the weight estimation device 1, the geometric quantity such as the total length can be known, and the CPU 7 can determine which of the data groups 3C to 3E the pig belongs to according to the value, and automatically change the formulas (1) to (3) to be applied.

[0042] In addition, the weight estimation device 1 may be provided with an interface such as a USB, allowing the CPU 7 to be accessed from an external personal computer, etc., so that a desired regression line can be written or changed in the memory unit 6 at any time (such as when shipped from the factory or after use by the user).

[0043] In the above example, the regression line is changed according to the age in months of the pig whose weight is to be estimated. On the other hand, the regression line can also be changed according to the type of pig. For example, there are various types of pigs, such as Landrace, Large Yorkshire, Duroc, Berkshire, Hampshire, and Medium Yorkshire, and the body shapes of each type are different. Therefore, a corresponding regression line may be obtained in advance for each of these pig types, and the user may select the type using a selection button provided on the weight estimation device 1.

[0044] In the above embodiment, an example was shown in which the regression line was expressed by a regression equation with one geometric quantity variable (surface area of ​​the lower body), but it may be expressed by a multiple regression equation with two or more geometric quantity variables. In addition, the regression equation is not limited to a linear equation, but may be a multi-order equation, or may be other functions (combinations of exponential functions, log functions, trigonometric functions, etc.).

[0045] Furthermore, even if they are the same breed, the growth of pigs will vary depending on the feed and environment. In this case, there are many different growth conditions, but in general, pigs are raised in the same pig house from piglets to adult pigs, so the growth of pigs will vary depending on the pig house.

[0046] Fig. 8 shows data in which the body lengths L of pigs raised in a first pigpen are plotted in correspondence with their measured weights W, and the regression line of the data is also superimposed. Fig. 9 shows data in which the body lengths L of pigs of the same age raised in a second pigpen different from the first pigpen are plotted in correspondence with their measured weights W, and the regression line of the data is also superimposed. Here, an example is shown in which weight is estimated using body length L as a geometric quantity.

[0047] For the data shown in Figure 8, the regression line is expressed as W = 0.1 L - 8 [Equation 4]. On the other hand, for the data shown in Figure 9, the regression line is expressed as W = 0.2 L - 132 [Equation 5]. Therefore, it can be seen that the regression lines used to estimate body weight are different even for pigs of the same breed and age.

[0048] Therefore, the body weight estimation device 1 is provided with a selection button for each pigpen, so that the user can select which pigpen the pigs in have been imaged by using the selection button. After that, by imaging the pigs using the body weight estimation device 1, the CPU 7 can select either the regression line of equation (4) or equation (5) that corresponds to the operated selection button, and can accurately estimate the body weight from the body length L.

[0049] Another factor that determines the regression line is the size of the pig. Here, size can refer to the size of a pig's parts, such as the thickness of its legs, or the overall size, such as its body length and height. Even in the same pig house, the regression line can change if the type of feed or amount of feed is changed. The regression line can also change if the pig house environment (temperature and stress factors) is changed.

[0050] Therefore, the characteristics that determine the regression line include the size, age, and breed of the pig, and the pig house in which it is raised, and further characteristics that determine the regression line include the type of feed, feeding policy, and pig house environment (temperature, humidity, ventilation, etc.).

[0051] The present invention is not limited to the above embodiment. For example, in imaging by the body weight estimation device 1, since the shutter speed is about 1 / 60, imaging is continuously performed to obtain a group of three-dimensional information, and the group of three-dimensional information is averaged to obtain the body weight W, or the surface area A of the pig's lower body obtained from each group of three-dimensional information is averaged to obtain one body weight W, or a plurality of obtained body weights W are averaged to use the average as the estimated body weight.

[0052] In the above embodiment, a pig is taken as an example of an animal whose weight is to be estimated, but the weight estimation device 1 according to the present embodiment can estimate the weight of a cow, a horse, etc. In such a case, since cows and horses have different body shapes from pigs, it is preferable to use different geometric quantities as parameters. For this reason, it is preferable to capture an image of a horse from behind as well as from the side to obtain a group of three-dimensional information.

[0053] It is preferable that the weight estimation device further includes a display device having a screen for displaying an image of a subject including an animal based on a signal output from the information acquisition device, and a selection device for selecting an animal displayed on the screen as a target for weight estimation, and the information acquisition device acquires the three-dimensional information of the animal selected by the selection device.

[0054] In the body weight estimation device, it is preferable that the selection device displays an identification mark on a screen of the display device, and selects an animal displayed superimposed on the identification mark as the body weight estimation target.

[0055] In the body weight estimation device, the identification mark may be provided in plurality.

[0056] In the body weight estimation device, it is preferable that the arithmetic device does not use the three-dimensional information for subjects displayed on the screen other than the animal selected by the selection device.

[0057] In the weight estimation device, it is preferable that the information acquisition device is a one-shot type 3D scanner that takes multiple shots within a predetermined time, and the calculation device calculates the geometric quantities of the animal's external shape based on average three-dimensional information obtained by averaging multiple pieces of three-dimensional information obtained by the multiple shots.

[0058] In the weight estimation device, it is preferable that the information acquisition device is a one-shot type 3D scanner that takes multiple shots within a predetermined time, the calculation device calculates the geometric quantities of the external shapes of the multiple animals based on the multiple three-dimensional information obtained from the multiple shots, and the weighing device estimates the weight of the imaged animal based on an average geometric quantity obtained by averaging the geometric quantities of the external shapes of the animals.

[0059] In the body weight estimation device, it is preferable that the information acquisition device is a one-shot type 3D scanner that takes multiple shots within a predetermined time, the calculation device calculates the geometric quantities of the external shapes of the multiple animals based on the multiple three-dimensional information obtained from the multiple shots, and the weighing device calculates multiple body weights of the imaged animals based on the geometric quantities of the external shapes of the multiple animals, and the average of these body weights is the estimated body weight. [Explanation of symbols]

[0060] 1. Weight Estimation Device 2. Housing 3 Control section 3a Power switch 3b Release button 4 Display 5. Imaging unit 6 Memory section

Claims

1. A weight estimation device for estimating the weight of an animal, comprising: An information acquisition device that is a one-shot type 3D scanner that captures an image of the animal and acquires three-dimensional information for each of a plurality of points; A calculation device that calculates a geometric quantity of the outer shape of the animal based on average three-dimensional information obtained by averaging a plurality of pieces of three-dimensional information acquired by the one-shot type 3D scanner performing a plurality of shots within a predetermined time; a weighing device that estimates a weight of the imaged animal based on the geometric quantity of the outer shape of the animal calculated by the calculation device; A memory unit that stores a regression equation that represents a correlation between the geometric quantity of the outer shape of the animal and the weight of the animal, the weighing device estimates the weight of the animal by substituting the geometric quantity of the external shape of the animal calculated by the calculation device into the regression equation; The memory unit stores a plurality of regression equations that represent a plurality of different correlations according to characteristics of an animal to be estimated, The weighing device selects one of the regression equations based on the geometric quantities of the animal's external shape calculated by the calculation device, reads it from the memory unit, and estimates the animal's weight by substituting the geometric quantities of the animal's external shape calculated by the calculation device.

2. A weight estimation device for estimating the weight of an animal, comprising: An information acquisition device that is a one-shot type 3D scanner that captures an image of the animal and acquires three-dimensional information for each of a plurality of points; A calculation device that calculates geometric quantities of the outer shapes of the animals based on a plurality of pieces of three-dimensional information acquired by the one-shot type 3D scanner performing a plurality of shots within a predetermined time; a weighing device that estimates a weight of the imaged animal based on an average geometric quantity obtained by averaging geometric quantities of a plurality of the animal's external shapes; A memory unit that stores a regression equation that represents a correlation between the geometric quantity of the outer shape of the animal and the weight of the animal, the weighing device estimates the weight of the animal by substituting the geometric quantity of the external shape of the animal calculated by the calculation device into the regression equation; The memory unit stores a plurality of regression equations that represent a plurality of different correlations according to characteristics of an animal to be estimated, The weighing device selects one of the regression equations based on the geometric quantities of the animal's external shape calculated by the calculation device, reads it from the memory unit, and estimates the animal's weight by substituting the geometric quantities of the animal's external shape calculated by the calculation device.

3. A weight estimation device for estimating the weight of an animal, comprising: An information acquisition device that is a one-shot type 3D scanner that captures an image of the animal and acquires three-dimensional information for each of a plurality of points; A calculation device that calculates geometric quantities of the outer shapes of the animals based on a plurality of pieces of three-dimensional information acquired by the one-shot type 3D scanner performing a plurality of shots within a predetermined time; a weighing device that calculates a plurality of weights of the imaged animals based on geometric quantities of the external shapes of the animals, and averages the weights to obtain an estimated weight; A memory unit that stores a regression equation that represents a correlation between the geometric quantity of the outer shape of the animal and the weight of the animal, the weighing device estimates the weight of the animal by substituting the geometric quantity of the external shape of the animal calculated by the calculation device into the regression equation; The memory unit stores a plurality of regression equations that represent a plurality of different correlations according to characteristics of an animal to be estimated, The weighing device selects one of the regression equations based on the geometric quantities of the animal's external shape calculated by the calculation device, reads it from the memory unit, and estimates the animal's weight by substituting the geometric quantities of the animal's external shape calculated by the calculation device.

4. 4. The weight estimation device according to claim 1, wherein the characteristic of the animal is a size of the animal, an age of the animal in months, or a pigpen in which the animal is kept.

Citation Information

Patent Citations

  • Apparatus for measuring weight of animal

    JP2002286421A

  • Estimation of physical parameters using three-dimensional representation

    JP2012519277A

  • Cow body diagnostic system and cow body diagnostic method

    JP2016059300A