Method for estimating feed components, method for determining blending amount of multiple types of feed, and computer program

Abstract

To provide a technology that can estimate feed components with a non-destructive method in a short period of time, or to provide a technology to appropriately blend feed according to feed components.SOLUTION: A method for estimating feed components comprises: (a) imaging each feed of the same type without pulverizing the feed for one or more types of feed with a spectral camera that can receive light with multiple pixels arranged in a two-dimensional array, comprising a spectral filter and an image sensor, and thereby obtaining a spectral reflectance from the light reception results at each of the multiple pixels, to obtain spectral reflectance information obtained according to the spectral reflectance for each of the one or more types of feed; (b) estimating the amount of the plurality of components for each of the one or more types of feed using a component estimation model that inputs the spectral reflectance information obtained from the light reception results of at least some of the multiple pixels and outputs the amount of the plurality of components; and (c) outputting the component amounts of the plurality of components.SELECTED DRAWING: Figure 5

Description

[Technical field]

[0001] The present disclosure relates to a method for estimating feed ingredients, a method for determining the blending amounts of multiple types of feed, and a computer program. [Background technology]

[0002] Patent Document 1 discloses a technology in which a computer automatically estimates the weight of an individual animal from image data of the individual animal, and the individual animal is guided to one of two feeding areas depending on the result of comparing the estimated weight with an evaluation standard weight. The two feeding areas are provided with food of different nutritional value. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2007-175050 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, the conventional technology does not take into consideration the variation in the components of the feed or the method of blending the feed. In reality, the components of the feed vary, so it is desirable to accurately measure or estimate the components of the feed. In general, there are two methods for measuring the components of feed: a method using chemicals and a method using spectroscopic information. The former method requires equipment, chemicals, and other specialized knowledge, so it is not a method that anyone who blends feed can perform. On the other hand, in the latter component analysis using spectroscopic information, since spectroscopic information can only be obtained at one point on the feed, it was necessary to eliminate the variation in the amount of components for each part, such as leaves and stems. In order to eliminate this variation, it was necessary to perform a crushing process before measurement, which caused the problem that the results could not be obtained immediately at the timing of feeding.

[0005] For these reasons, conventional ingredient analysis methods are destructive and require specialized knowledge and technology, and they cannot be performed on-site in real time.In addition, there is a possibility that the quality of the feed may change during ingredient analysis, which makes it difficult to properly design the feed mix.

[0006] Therefore, there is a need for a technology that allows anyone at the feeding site to quickly and non-destructively determine the composition of feed. There is also a need for a technology that allows appropriate feed compounding according to the composition of the feed. [Means for solving the problem]

[0007] According to a first embodiment of the present disclosure, there is provided a method for estimating ingredients of feed, which includes the steps of: (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at a plurality of pixels arranged two-dimensionally, capturing an image of one or more types of feed for each type of feed without crushing the feed to obtain a spectral reflectance from the light reception results at each of the plurality of pixels, and determining spectral reflectance information obtained according to the spectral reflectance for each of the one or more types of feed, (b) estimating the amounts of the plurality of ingredients for each of the one or more types of feed using a ingredient estimation model that receives the spectral reflectance information obtained from the light reception results at at least some of the plurality of pixels as an input and outputs the amounts of the plurality of ingredients, and (c) outputting the amounts of the plurality of ingredients.

[0008] According to a second aspect of the present disclosure, there is provided a method for determining the mixing amounts of multiple types of feed for multiple reared individuals, which includes the steps of: (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at multiple pixels arranged two-dimensionally, capturing images of each type of feed without crushing the feed to obtain spectral reflectance from the light reception results at each of the multiple pixels, and determining spectral reflectance information obtained according to the spectral reflectance for each of the multiple types of feed; and (b) estimating the amounts of the multiple components for each of the multiple types of feed using a component estimation model that receives the spectral reflectance information obtained from the light reception results at at least some of the multiple pixels as an input and outputs the amounts of the multiple components. (c) acquiring individual rearing information regarding the rearing condition of each of the plurality of reared individuals; (d) determining the required amounts of the plurality of components for each of the plurality of reared individuals using a required component estimation model that takes the individual rearing information as input and outputs the required amounts of the plurality of components; and (e) determining the mixing amounts of the plurality of types of feed for each of the plurality of reared individuals from the amounts of the plurality of components for each of the plurality of types of feed estimated in (b) and the required amounts of the plurality of components for each of the plurality of reared individuals determined in (d).

[0009] According to a third aspect of the present disclosure, there is provided a computer program for causing a processor to execute a process for estimating ingredients of a feed. This computer program causes the processor to execute the following processes: (a) using a spectroscopic camera including a spectral filter and an image sensor capable of receiving light at a plurality of pixels arranged two-dimensionally, for one or more types of feed, to obtain a spectral reflectance from the light reception results at each of the plurality of pixels by capturing an image of each type of feed without crushing the feed, and to obtain spectral reflectance information obtained according to the spectral reflectance for each of the one or more types of feed, (b) using a component estimation model that receives the spectral reflectance information obtained from the light reception results at at least a portion of the plurality of pixels as an input and outputs the amounts of the plurality of components, to estimate the amounts of the plurality of components for each of the one or more types of feed, and (c) output the amounts of the plurality of components.

[0010] According to a fourth aspect of the present disclosure, there is provided a computer program for causing a processor to execute a process of determining the mixing amounts of multiple types of feed for multiple reared individuals. This computer program includes the following steps: (a) using a spectroscopic camera including a spectral filter and an image sensor capable of receiving light at multiple pixels arranged two-dimensionally, the multiple types of feed are imaged for each type of feed without crushing the feed, to obtain spectral reflectance from the light reception results at each of the multiple pixels, and to obtain spectral reflectance information obtained according to the spectral reflectance for each of the multiple types of feed; and (b) using a component estimation model that receives the spectral reflectance information obtained from the light reception results at at least some of the multiple pixels as an input and outputs the component amounts of multiple components, to estimate the component amounts of the multiple components for each of the multiple types of feed. The processor is caused to execute the following steps: (c) a process for acquiring individual rearing information regarding the rearing condition of each of the plurality of reared individuals for each of the plurality of reared individuals; (d) a process for determining the required amounts of the plurality of components for each of the plurality of reared individuals using a required component estimation model that takes the individual rearing information as input and outputs the required amounts of the plurality of components; and (e) a process for determining the mixing amounts of the plurality of types of feed for each of the plurality of reared individuals from the amounts of the plurality of components for each of the plurality of types of feed estimated in process (b) and the required amounts of the plurality of components for each of the plurality of reared individuals determined in process (d). [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 is an explanatory diagram showing the configuration of a feeding system according to an embodiment. [Diagram 2] FIG. 1 is a block diagram showing a configuration of an information processing device. [Diagram 3] FIG. 2 is a block diagram showing the internal configurations of a machine learning processing unit, a feed component amount estimation unit, and a feed blending processing unit. [Figure 4] 11 is a flowchart showing the procedure of machine learning processing. [Diagram 5]11 is a flowchart showing the processing steps of a component amount estimation process and a blend amount determination process. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] 1 is an explanatory diagram showing the configuration of a feeding system according to an embodiment. The feeding system includes a livestock house 100, an automatic feeding device 200, an information processing device 300, and a spectroscopic camera 400.

[0013] The livestock shed 100 has a plurality of barns 110. One animal BA is housed in each barn 110. The animal BA is, for example, a cow or a pig. The barn 110 is provided with a feeding trough 111 for mixed feed.

[0014] The automatic feeding device 200 supplies mixed feed that is appropriately mixed according to the breeding condition of the reared individuals BA to the feeding trough 111 individually. The automatic feeding device 200 has a plurality of feed tanks 210 and a mixing and distributing device 220. The plurality of feed tanks 210 each contain different feeds FS1 to FSn. Here, the subscript n is an integer of 2 or more. The mixing and distributing device 220 prepares mixed feed by mixing the feeds FS1 to FSn according to the breeding condition of each reared individual BA according to the feed blending information provided by the information processing device 300, and supplies the mixed feed to the feeding trough 111.

[0015] The spectroscopic camera 400 is for measuring the spectral intensity distribution of the feed FSj. Here, the subscript j is an integer from 1 to n. The spectroscopic camera 400 includes a spectroscopic filter 401 and an image pickup element 402. As shown in an enlarged view at the bottom of FIG. 1, in this embodiment, the image pickup element 402 is a two-dimensional element in which a plurality of light receiving elements arranged two-dimensionally function as a plurality of pixels Px, and can receive light at these pixels Px. However, the spectroscopic camera 400 may be configured to receive light at a plurality of two-dimensionally arranged pixels Px by scanning a one-dimensional element in which light receiving elements are linearly arranged. The spectroscopic filter 401 is a wavelength-variable transmission filter that has a narrow half-width of the transmission wavelength and can change the transmission wavelength. In addition, it is preferable that the spectroscopic filter 401 is one that passes only light in the near-infrared region. The near-infrared region typically means a wavelength range of about 700 to about 2500 nm. In this embodiment, an optical filter capable of changing the transmission wavelength in the wavelength range of 680 to 950 nm in the near infrared region is used as the spectral filter 401. Such a spectral filter 401 can be realized by using, for example, a Fabry-Perot type spectral filter.

[0016] When photographing with the spectroscopic camera 400, the feed FSj is contained in a container 410 such as a petri dish, and is obliquely illuminated from the left and right by the illumination light source 420. For example, a halogen light source is used as the illumination light source 420. The spectral reflectance of the feed FSj is a value obtained by dividing the spectral intensity I(λ) obtained by measuring the feed FSj by the reference spectral intensity I0(λ) obtained by measuring a standard white plate. Since the spectral intensities I(λ) and I0(λ) are measured for each of a plurality of pixels Px arranged two-dimensionally, the spectral reflectance is calculated using the spectral intensities I(λ) and I0(λ) obtained at the same pixel position (u, v). Note that the spectral intensities I(λ) used to obtain the components of the feed FSj do not need to be obtained from all the pixels Px of the spectroscopic camera 400, and only the spectral intensities obtained from some of the pixels Px may be used.

[0017] Incidentally, a spectroscopic measuring device that is composed of a spectroscopic filter and a photoelectric conversion element such as a photodiode is known. However, in such a spectroscopic measuring device, the area that can be spectroscopically measured with high accuracy is limited, so it is difficult to spectroscopically measure a large measurement sample with a side of about several tens of centimeters in a short time. On the other hand, the spectroscopic camera 400 of this embodiment includes a spectroscopic filter 401 and an image pickup element 402, and is configured to be able to receive light with a plurality of pixels Px arranged two-dimensionally, so that it has the advantage of being able to spectroscopically measure a large measurement sample in a short time. In addition, by using the spectroscopic camera 400 of this embodiment, it is possible to obtain a feed image having intensity information for each wavelength, so that the influence of unevenness in the feed can be reduced, and there is an advantage that pre-processing such as crushing and drying the feed before spectroscopic measurement is not required.

[0018] The information processing device 300 executes a process of estimating the amounts of a plurality of ingredients contained in the feed FSj using the spectral reflectance of each feed FSj. The information processing device 300 further executes a process of determining the blending amount of the feed suitable for the reared individual BA according to the amounts of the ingredients of the feed FSj and the individual rearing information of each reared individual BA.

[0019] 2 is a block diagram showing the configuration of the information processing device 300. The information processing device 300 has a processor 310, a memory 320, an interface circuit 330, and an input device 340 and a display device 350 connected to the interface circuit 330. A spectroscopic camera 400 is also connected to the interface circuit 330. The processor 310 not only has a function of executing the processes described in detail below, but also has a function of displaying on the display device 350 data obtained by the processes and data generated in the course of the processes.

[0020] The processor 310 has the functions of a machine learning processing unit 510 that performs learning of a machine learning model, a feed component amount estimation unit 520 that estimates the amounts of multiple components contained in the feed FSj, and a feed blending processing unit 530 that determines the blending amount of feed suitable for the reared individual BA. These functions are respectively realized by the processor 310 executing a computer program stored in the memory 320. However, some of these functions may be realized by a hardware circuit. The term processor in the present disclosure includes such a hardware circuit. In addition, one or more processors that perform various processes may be processors included in one or more remote computers connected via a network.

[0021] The memory 320 stores a component estimation model 610, a required component estimation model 620, a reared individual database 630, and a feed database 640. The component estimation model 610 is a machine learning model that receives spectral reflectance information of the feed FSj as input and outputs the component amounts of multiple components contained in the feed FSj. The required component estimation model 620 is a machine learning model that receives individual rearing information related to the rearing state of the reared individual BA as input and outputs the required component amounts of multiple components. The reared individual database 630 is a database in which the individual ID and rearing history of each of the multiple reared individuals BA are registered. The feed database 640 is a database in which the component amounts of multiple components are registered for multiple types of feed FSj.

[0022] FIG. 3 is a block diagram showing the internal configuration of the machine learning processing unit 510, the feed component amount estimating unit 520, and the feed blending processing unit 530.

[0023] The machine learning processing unit 510 includes a preprocessing unit 511, a feed component amount input unit 512, and a learning execution unit 513. The machine learning processing unit 510 executes machine learning of the component estimation model 610 using a plurality of learning feeds LFSi. Here, the subscript i is an integer from 1 to M, and M is an integer of 2 or more. As described above, the component estimation model 610 is a machine learning model that inputs the spectral reflectance information of the feed and outputs the component amounts of a plurality of components contained in the feed. The component estimation model 610 can be configured as a model using, for example, linear regression such as ridge regression or a decision tree such as random forest. However, the component estimation model 610 may be configured using other types of models such as a neural network or a support vector machine.

[0024] The preprocessing unit 511 creates spectral reflectance information by performing preprocessing on the spectral reflectance of the learning feed LFSi obtained by the spectroscopic camera 400. As the preprocessing, for example, any one of the calculations of the following formulas (1) to (3) can be used.

number

[0025] The spectral reflectance R(λ) is the result of dividing the spectral intensity I(λ) by the reference spectral intensity I0(λ), and is calculated for each pixel Px. In order to remove the influence of measurement noise, m spectral reflectances may be calculated from the light reception results of Np pixels Px to generate m pieces of spectral reflectance information. Here, Np is the total number of pixels Px arranged in a two-dimensional array, and m is an integer not less than 2 and not more than Np. As a method for calculating the m spectral reflectances, for example, a pixel region consisting of Np pixels is divided into m small regions, and the spectral intensity I and the reference spectral intensity I0 are respectively averaged in each small region, and the averaged spectral intensity is divided by the averaged reference spectral intensity to calculate the m spectral reflectances. When each small region is a rectangular region including M pixels, the averaging means taking the average of M pixels. Here, M is an integer not less than 2 and not more than Np. After calculating the spectral reflectance for each Np pixel, the spectral reflectances may be averaged in each small region. As another method for obtaining m spectral reflectances, for example, a method can be used in which an error between an average spectral reflectance obtained by averaging Np spectral reflectances for each wavelength and each individual spectral reflectance is obtained, and a spectral reflectance with an error value equal to or less than a predetermined threshold is selected. For example, a root mean square error can be used as the error. In this way, it is possible to exclude spectral reflectances with low reliability. Note that the spectral reflectance information may be generated using one representative spectral reflectance that represents the selected m spectral reflectances. The representative spectral reflectance may be an average or a linear sum of the m spectral reflectances. In addition, as another method for removing the influence of measurement noise, M spectral reflectances may be calculated from Np pixels, and m spectral reflectances may be selected from the M spectral reflectances. Here, M is an integer equal to or greater than 2 and less than Np, and m is an integer equal to or greater than 2 and M. By performing a calculation according to any of the above-mentioned formulas on the m spectral reflectances obtained by any of these methods, it is possible to obtain m spectral reflectance information. This process corresponds to a process of determining m pieces of spectral reflectance information using light reception results at m pixels selected from a plurality of pixels.

[0026] In the above formulas (1) to (3), the base of the logarithm is set to 10, but the base of the logarithm may be a value other than 10. As can be understood from these examples, it is preferable that the spectral reflectance information is information calculated from the logarithm of the spectral reflectance R(λ). According to an experiment by the inventor of the present disclosure, it has been found that in particular, when the second derivative U(λ) of the logarithm of the spectral reflectance R(λ) is used as the spectral reflectance information, it is preferable in that the estimated value of the component amount of the feed by the component estimation model 610 becomes more accurate. The reason for taking the logarithm of the spectral reflectance R(λ) is that in the light absorption phenomenon caused by the target component, the logarithm of the spectral reflectance R(λ) is proportional to the concentration of the target component. In addition, the reason for taking the second derivative is that which wavelength is more strongly absorbed differs depending on the target component, and this appears as a peak in the spectral reflectance distribution. In other words, by taking the second derivative of the logarithm of the spectral reflectance R(λ), it is possible to extract the peak of the spectral reflectance distribution with high sensitivity.

[0027] The feed ingredient amount input unit 512 receives input of the ingredient amounts of the multiple ingredients for each of the multiple learning feeds LFSi. For each learning feed LFSi, the ingredient amounts of the multiple ingredients are accurately measured in advance using a chemical ingredient analyzer. The multiple ingredients are ingredients that are highly important as feed for the reared individual BA, and for example, multiple ingredients including moisture, crude protein, crude fat, neutral detergent fiber, and acid detergent fiber are used. It is preferable that the multiple ingredients contain moisture in particular. According to the ingredient estimation process described later, anyone can estimate the ingredients of the feed non-destructively in a short time, so there is no need to worry about the deterioration of the feed that occurred during the conventional work ingredient analysis, and multiple ingredients including moisture can be accurately and quickly estimated at the feeding site.

[0028] The learning execution unit 513 executes machine learning of the component estimation model 610 using the spectral reflectance information on the multiple learning feeds LFSi and the component amounts of the multiple components as learning data.

[0029] The feed component amount estimation unit 520 includes a preprocessing unit 521, an estimation execution unit 522, and a component estimation model 610. The component estimation model 610 is stored in the memory 320 after the machine learning processing unit 510 has completed machine learning. The preprocessing unit 521 executes the same preprocessing as the preprocessing unit 511 of the machine learning processing unit 510. The estimation execution unit 522 executes a process of estimating the component amounts of multiple components for each of multiple types of feed FSj by inputting the spectral reflectance information of the feed FSj to the machine-learned component estimation model 610. These feeds FSj are feeds that are actually stored in multiple feed tanks 210 of the automatic feeding device 200.

[0030] The feed blending processing unit 530 includes a breeding information acquisition unit 531 , a required component amount determination unit 532 , a feed blending amount determination unit 533 , a required component estimation model 620 , a breeding individual database 630 , and a feed database 640 .

[0031] The breeding information acquisition unit 531 acquires individual breeding information on the breeding state of each of the multiple breeding individuals BA from the breeding individual database 630. As the individual breeding information, for example, two or more pieces of information such as individual ID, age in months, weight, feeding history, and BCS (body condition score) can be used. Note that the individual breeding information may be acquired using the measurement results from various sensors such as a weight scale and a stereo camera without using the breeding individual database 630.

[0032] The required component amount determination unit 532 uses the required component estimation model 620 to determine the required component amounts of the multiple components for each of the multiple reared individuals BA. The required component estimation model 620 is a model that receives individual rearing information as an input and outputs the required component amounts of the multiple components. The required component estimation model 620 may be configured to receive a value index related to the value of the reared individual BA as a livestock, in addition to the individual rearing information of the reared individual BA. As this value index, the milk components of dairy cows and the meat quality of beef cattle can be used. The required component estimation model 620 may be configured as a machine learning model such as a neural network, or may be configured as a function or a lookup table. It is assumed that the learning of the required component estimation model 620 has been performed in advance.

[0033] The feed blending amount determination unit 533 determines the blending amounts of multiple types of feed FSj for each reared individual BA from the component amounts of multiple components for each of the feeds FSj estimated by the feed component amount estimation unit 520 and the required component amounts for the reared individual BA determined by the required component amount determination unit 532. Feed blending information indicating the blending amounts of multiple types of feed FSj is supplied to the automatic feeding device 200.

[0034] 4 is a flowchart showing the procedure of the machine learning process. This machine learning process is performed by the machine learning processing unit 510.

[0035] In step S110, the feed component amount input unit 512 acquires the component amounts of multiple components for multiple learning feeds LFSi. The component amounts here are, for example, the component amounts per unit mass. As described above, the component amounts of multiple components contained in the learning feed LFSi are accurately measured in advance. In step S120, the machine learning processing unit 510 acquires the spectral reflectance R(λ) for each of the multiple learning feeds LFSi using the spectroscopic camera 400.

[0036] In step S130, the preprocessing unit 511 generates spectral reflectance information by performing preprocessing on the spectral reflectance R(λ). As described above, the spectral reflectance R(λ) is a result of dividing the spectral intensity I(λ) by the reference spectral intensity I0(λ), and is calculated for each pixel Px. Note that the spectral reflectance and the spectral reflectance information may be generated using only a part of the pixels Px. For example, in order to remove the influence of measurement noise, m spectral reflectances may be obtained from the light reception results of Np pixels Px, and m pieces of spectral reflectance information may be generated. Furthermore, one representative spectral reflectance representing the m spectral reflectances may be used to generate one piece of spectral reflectance information. As described above, it is preferable that the spectral reflectance information is calculated from the logarithm of the spectral reflectance R(λ), and it is particularly preferable that the spectral reflectance information is calculated by second-order differentiation of the logarithm of the spectral reflectance.

[0037] In step S140, the learning execution unit 513 uses learning data including the component amounts and spectral reflectance information of the learning feed LFSi to execute machine learning of the component estimation model 610. When the learning is completed, the learned component estimation model 610 is stored in the memory 320.

[0038] In this way, the machine learning process of this embodiment can effectively execute machine learning of the component estimation model 610 that estimates the components of the feed. In particular, since information calculated from the logarithm of the spectral reflectance R(λ) is used as the spectral reflectance information, it is possible to improve the estimation accuracy of the component estimation model 610.

[0039] 5 is a flow chart showing the process steps of the ingredient amount estimation process and the blending amount determination process. These processes are performed by the feed ingredient amount estimation unit 520 and the feed blending process unit 530.

[0040] In step S210, the feed component amount estimation unit 520 acquires the spectral reflectance R(λ) for each of the multiple types of feed FSj using the spectroscopic camera 400. The spectroscopic measurement of the feed FSj is performed without crushing the feed FSj. It is also preferable that the spectroscopic measurement of the feed FSj is performed without drying the feed FSj. In this embodiment, the spectroscopic measurement of the feed FSj is performed without crushing or drying the feed FSj, so that the feed FSj that has been the subject of the spectroscopic measurement can be used as it is as food for the reared individual BA.

[0041] In step S220, the preprocessing unit 521 generates spectral reflectance information by performing preprocessing on the spectral reflectance R(λ). This preprocessing is the same as the preprocessing in step S130.

[0042] In step S230, the estimation execution unit 522 estimates the amounts of multiple components for each feed FSj using the trained component estimation model 610. That is, the estimation execution unit 522 inputs the spectral reflectance information of each feed FSj to the component estimation model 610, and obtains the amounts of multiple components output from the component estimation model 610. The component amounts estimated here are, for example, the component amounts per unit mass.

[0043] When m pieces of spectral reflectance information calculated from m pieces of light receiving results are used as the spectral reflectance information of one type of feed FSj, it is preferable to determine, for each component, a representative value of m estimated component amounts estimated from m pieces of spectral reflectance information as the estimated value of the component amount of that component. Here, m is an integer of 2 or more and Np, where Np is the total number of pixels Px. The representative value of the m estimated component amounts can be the average value, maximum value, minimum value, etc. of the m estimated component amounts. However, if the average value of the m estimated component amounts is used as the representative value, the estimation accuracy can be improved.

[0044] In step S240, the estimation execution unit 522 outputs estimated values ​​of the component amounts of the multiple components for each feed FSj. In this embodiment, the estimation execution unit 522 registers the estimated values ​​of the component amounts of the multiple components in the feed database 640. Note that in step S240, other indicators such as the total amount of digestible nutrients calculated from the component amounts of the multiple components may be output. Also, the estimated values ​​of the component amounts of the multiple components may be output to an output device such as the display device 350 or a printer.

[0045] In step S240, the estimation execution unit 522 may further output a distribution of the component amount in an image region including Np pixels Px. In the lower part of FIG. 1, a component amount distribution CDi displayed on the display device 350 is illustrated as an example of this output. The component amount distribution CDi is an image showing the component amount of the i-th component in an image region including Np pixels Px. In the example of FIG. 1, the component amount is expressed by pixel concentration, but the component amount may be expressed by a numerical value. Such a component amount distribution CDi does not need to be created for an image region including all Np pixels Px, and may be created for an image region including at least a part of the Np pixels Px. In addition, it is preferable that the component amount distribution CDi is created for at least one component among the multiple components. By outputting such a component amount distribution CDi, the operator can know the distribution of the component amount in the feed FSj.

[0046] The above processing of steps S210 to S240 corresponds to processing for estimating the components of the feed. In the above example, the component amounts were estimated for each of the multiple types of feed FSj, but the method disclosed herein is also applicable to the case where the component amounts are estimated for one or more types of feed. However, in the processing from step S250 onwards, the estimated values ​​of the component amounts for the multiple types of feed FSj are used to determine the feed blend amount for each reared individual BA.

[0047] In step S250, the breeding information acquisition unit 531 acquires individual breeding information for one raised individual BA. The individual breeding information can be acquired by referring to the raised individual database 630 using the individual ID of the raised individual BA. The individual breeding information includes, for example, two or more pieces of information such as the individual ID, age in months, weight, feeding history, and BCS (body condition score). Note that some or all of the individual breeding information may be acquired from each raised individual BA using various sensors. The following processing of steps S260 to S280 is executed for one raised individual BA selected in step S250.

[0048] In step S260, the required component amount determination unit 532 uses the required component estimation model 620 to determine the required component amounts of multiple components for the reared individual BA. That is, the required component amount determination unit 532 inputs the individual rearing information of the reared individual BA to the required component estimation model 620, and obtains the required component amounts of multiple components output from the required component estimation model 620. The required component amounts are, for example, the weights of the components. Note that, when the required component estimation model 620 is configured to input a value index related to the value of the reared individual BA as a livestock in addition to the individual rearing information of the reared individual BA, the required component amount determination unit 532 may accept the input of the value index. Alternatively, the value index may be set in advance before starting the processing of FIG. 5.

[0049] In step S270, the feed blending amount determination unit 533 determines the blending amounts of multiple types of feed FSj for the reared individual BA. That is, the feed blending amount determination unit 533 determines the blending amounts of multiple types of feed FSj so as to achieve the required component amounts of the multiple components determined in step S260. At this time, the values ​​registered in the feed database 640 in step S230 are used as the amounts of the multiple components contained in each feed FSj. The blending amounts of the multiple types of feed FSj are determined so as to minimize, for example, an evaluation value E given by the following formula.

number

[0050] By determining the blending amounts Mj of multiple types of feed FSj so that the evaluation value E given by the above formula (4) is minimized, the required component amounts Ci of multiple components i can be achieved.

[0051] As shown in the following formula (5), a necessary condition may be set that the error in the component amount for each component i is equal to or smaller than the allowable component amount error σi.

number

[0052] Instead of or in addition to the condition in formula (5) above, a necessary condition may be set that each component i contains an amount of component equal to or greater than the required component amount Ci, as shown in formula (6) below.

number

[0053] In step S280, the feed blending processing unit 530 supplies feed blending information indicating the blending amounts Mj of the multiple types of feed FSj determined in step S270 to the automatic feeding device 200, and the mixing and distributing device 220 blends and feeds the feed FSj in response to this.

[0054] It should be noted that the blending and feeding of the feed FSj may be performed manually without using the automatic feeding device 200. In this case, the information processing device 300 may output feed blending information indicating the blending amount Mj of the feed FSj using an output device such as the display device 350 or a printer. Alternatively, the information processing device 300 may transmit the feed blending information to another information processing device.

[0055] In step S290, it is determined whether the processing of steps S250 to S280 has been completed for all of the breeding individuals BA. If the processing of all of the breeding individuals BA has not been completed, the process returns to step S250, the next breeding individual BA is selected, and the processing of steps S250 to S280 is executed.

[0056] According to the component amount estimation process described above, anyone can estimate the components of the feed FSj in a short time and non-destructively. In addition, there is no need to worry about the deterioration of the feed that occurred during the conventional work component analysis, and feed design can be performed with higher accuracy. Furthermore, according to the above-mentioned mixture amount determination process, the mixture amounts of multiple types of feed FSj can be appropriately determined depending on the breeding condition of the breeding individual BA.

[0057] Other forms: The present disclosure is not limited to the above-mentioned embodiment, and can be realized in various forms without departing from the spirit of the present disclosure. For example, the present disclosure can be realized in the following aspects. The technical features in the above-mentioned embodiments corresponding to the technical features in each aspect described below can be appropriately replaced or combined in order to solve some or all of the problems of the present disclosure, or to achieve some or all of the effects of the present disclosure. Furthermore, if the technical feature is not described as essential in this specification, it can be appropriately deleted.

[0058] (1) According to a first embodiment of the present disclosure, there is provided a method for estimating ingredients of feed, which includes the steps of: (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at a plurality of pixels arranged two-dimensionally, capturing an image of one or more types of feed for each type of feed without crushing the feed to obtain a spectral reflectance from the light reception results at each of the plurality of pixels, and determining spectral reflectance information obtained according to the spectral reflectance for each of the one or more types of feed, (b) estimating the amounts of the ingredients of the ingredients of the one or more types of feed using a ingredient estimation model that receives the spectral reflectance information obtained from the light reception results at at least some of the plurality of pixels as an input and outputs the amounts of the ingredients of the ingredients, and (c) outputting the amounts of the ingredients of the ingredients. This method allows anyone to estimate the composition of feed in a short time and non-destructively. It also eliminates the risk of feed deterioration that occurred during conventional composition analysis, allowing for more accurate feed design.

[0059] (2) According to a second aspect of the present disclosure, a method for determining the amounts of multiple types of feed to be mixed with multiple reared individuals is provided, which includes the steps of: (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at multiple pixels arranged two-dimensionally, capturing images of each type of feed without crushing the feed to obtain spectral reflectance from the light reception results at each of the multiple pixels, and determining spectral reflectance information obtained according to the spectral reflectance for each of the multiple types of feed; and (b) estimating the amounts of multiple components for each of the multiple types of feed using a component estimation model that receives the spectral reflectance information obtained from the light reception results at at least some of the multiple pixels as an input and outputs the amounts of multiple components. (c) acquiring individual rearing information regarding the rearing condition of each of the plurality of reared individuals; (d) determining the required amounts of the plurality of components for each of the plurality of reared individuals using a required component estimation model that takes the individual rearing information as input and outputs the required amounts of the plurality of components; and (e) determining the mixing amounts of the plurality of types of feed for each of the plurality of reared individuals from the amounts of the plurality of components for each of the plurality of types of feed estimated in (b) and the required amounts of the plurality of components for each of the plurality of reared individuals determined in (d). According to this method, the mixing amounts of multiple types of feed can be appropriately determined depending on the breeding conditions of the animals being raised.

[0060] (3) In the above method, the step (a) may include a step of calculating m pieces of the spectral reflectance information using the light receiving results of m pixels selected from the plurality of pixels, where m is an integer equal to or greater than 2, and the step (b) may include a step of determining, as the component amounts of the plurality of components, representative values ​​of the m component amount estimates obtained from the component estimation model in accordance with each of the m pieces of spectral reflectance information. According to this method, the component amounts can be estimated with high accuracy using m pieces of spectral reflectance information.

[0061] (4) In the above method, the spectral reflectance information may be information calculated from a logarithm of the spectral reflectance. According to this method, the component amounts can be estimated with high accuracy using a component estimation model.

[0062] (5) In the above method, the spectral reflectance information may be information calculated by second-order differentiation of a logarithm of the spectral reflectance. According to this method, the component amounts can be estimated with higher accuracy using a component estimation model.

[0063] (6) In the above method, the spectral reflectance information may include information in a near-infrared region. According to this method, the component amounts can be estimated with high accuracy using a component estimation model.

[0064] (7) In the above method, the step (c) may include a step of outputting a distribution of the component amounts for at least a portion of the plurality of components in an image region that includes the at least a portion of the plurality of pixels. This method makes it possible to know the distribution of component amounts in the feed.

[0065] (8) In the above method, step (d) may involve determining the required component amounts of the plurality of components for each of the reared individuals by inputting a value index relating to the value of each feed individual as a livestock together with the individual rearing information into a required component estimation model. According to this method, the mixing amounts of multiple types of feed can be appropriately determined depending on the breeding condition and value index of the individual animals.

[0066] (9) According to a third aspect of the present disclosure, there is provided a computer program for causing a processor to execute a process for estimating ingredients of a feed. This computer program causes the processor to execute the following processes: (a) using a spectroscopic camera including a spectral filter and an image sensor capable of receiving light at a plurality of pixels arranged two-dimensionally, for one or more types of feed, to obtain a spectral reflectance from the light reception results at each of the plurality of pixels by capturing an image of each type of feed without crushing the feed, and to obtain spectral reflectance information obtained according to the spectral reflectance for each of the one or more types of feed, (b) using a component estimation model that receives the spectral reflectance information obtained from the light reception results at at least a portion of the plurality of pixels as an input and outputs the amounts of the plurality of components, to estimate the amounts of the plurality of components for each of the one or more types of feed, and (c) output the amounts of the plurality of components.

[0067] (10) According to a fourth aspect of the present disclosure, there is provided a computer program for causing a processor to execute a process of determining the mixing amounts of multiple types of feed for multiple reared individuals. This computer program includes: (a) a process of obtaining spectral reflectance from the light reception results at each of the multiple pixels by using a spectroscopic camera that includes a spectral filter and an image sensor and is capable of receiving light at multiple pixels arranged two-dimensionally for each type of feed, without crushing the feed, for each type of feed, and determining spectral reflectance information obtained according to the spectral reflectance for each of the multiple types of feed; and (b) a process of estimating the amounts of multiple components for each of the multiple types of feed using a component estimation model that receives the spectral reflectance information obtained from the light reception results at at least some of the multiple pixels as an input and outputs the amounts of multiple components. The processor is caused to execute the following steps: (c) a process for acquiring individual rearing information regarding the rearing condition of each of the plurality of reared individuals for each of the plurality of reared individuals; (d) a process for determining the required amounts of the plurality of components for each of the plurality of reared individuals using a required component estimation model that takes the individual rearing information as input and outputs the required amounts of the plurality of components; and (e) a process for determining the mixing amounts of the plurality of types of feed for each of the plurality of reared individuals from the amounts of the plurality of components for each of the plurality of types of feed estimated in process (b) and the required amounts of the plurality of components for each of the plurality of reared individuals determined in process (d).

[0068] The present disclosure may be realized in various forms other than those described above, such as an apparatus for carrying out the above-described method, a non-transitory storage medium having a computer program recorded thereon, and the like. [Explanation of symbols]

[0069] 100... livestock barn, 110... livestock pen, 111... feeding trough, 200... automatic feeding device, 210... feed tank, 220... mixing and distributing device, 300... information processing device, 310... processor, 320... memory, 330... interface circuit, 340... input device, 350... display device, 400... spectroscopic camera, 401... spectroscopic filter, 402... image sensor, 410... storage container, 420... illumination light source, 510... Machine learning processing unit, 511...pre-processing unit, 512...feed component amount input unit, 513...learning execution unit, 520...feed component amount estimation unit, 521...pre-processing unit, 522...estimation execution unit, 530...feed blending processing unit, 531...breeding information acquisition unit, 532...required component amount determination unit, 533...feed blending amount determination unit, 610...component estimation model, 620...required component estimation model, 630...breeding individual database, 640...feed database

Claims

1. A method for estimating the components of feed using a computer, comprising: (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at a plurality of pixels arranged two-dimensionally, to capture images of one or more types of feed for each type of feed without crushing the feed, thereby obtaining spectral reflectance from the light reception results at each of the plurality of pixels, and determining spectral reflectance information obtained in accordance with the spectral reflectance for each of the one or more types of feed; (b) estimating the amounts of the plurality of components for each of the one or more types of feed using a component estimation model that receives the spectral reflectance information obtained from the light reception results of at least some of the plurality of pixels as an input and outputs the amounts of the plurality of components; (c) outputting the component amounts of the plurality of components; A method comprising:

2. A method for determining the mixing amounts of multiple types of feed for multiple reared individuals by a computer, comprising: (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at a plurality of pixels arranged two-dimensionally, to capture images of the plurality of types of feed without crushing the feed for each type, thereby obtaining spectral reflectance from the light reception results at each of the plurality of pixels, and determining spectral reflectance information obtained according to the spectral reflectance for each of the plurality of types of feed; (b) estimating the amounts of the plurality of components for each of the plurality of types of feed using a component estimation model that receives the spectral reflectance information obtained from the light reception results of at least some of the plurality of pixels as an input and outputs the amounts of the plurality of components; (c) acquiring individual breeding information regarding the breeding status of each of the plurality of breeding individuals; (d) determining the required component amounts of the plurality of components for each of the plurality of reared individuals using a required component estimation model that receives the individual rearing information as input and outputs the required component amounts of the plurality of components; (e) determining the compounding amounts of the plurality of types of feed for each of the plurality of reared individuals from the component amounts of the plurality of components for each of the plurality of types of feed estimated in the step (b) and the required component amounts of the plurality of components for each of the plurality of reared individuals determined in the step (d); A method comprising:

3. 3. The method of claim 1 or 2, the step (a) includes a step of obtaining m pieces of the spectral reflectance information using the light reception results of m pixels selected from the plurality of pixels, where m is an integer equal to or greater than 2; the step (b) includes a step of determining, as the component amounts of the plurality of components, representative values of m component amount estimates obtained from the component estimation model in accordance with the m pieces of spectral reflectance information, respectively. method.

4. 3. The method of claim 1 or 2, The method, wherein the spectral reflectance information is information calculated from the logarithm of the spectral reflectance.

5. 5. The method of claim 4, The method, wherein the spectral reflectance information is information calculated by second-order differentiation of the logarithm of the spectral reflectance.

6. 3. The method of claim 1 or 2, The method, wherein the spectral reflectance information includes information in the near infrared region.

7. 10. The method of claim 1, The method, wherein step (c) includes a step of outputting, for at least some of the plurality of components, a distribution of the component amounts in an image region including the at least some of the plurality of pixels.

8. 3. The method of claim 2, The step (d) determines the required component amounts of the plurality of components for each individual raised by inputting a value index relating to the value of each individual feed animal as a livestock, together with the individual rearing information, into a required component estimation model.

9. A computer program that causes a processor to execute a process for estimating ingredients of feed, (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at a plurality of pixels arranged two-dimensionally, to capture images of one or more types of feed for each type of feed without crushing the feed, thereby obtaining spectral reflectance from the light reception results at each of the plurality of pixels, and determining spectral reflectance information obtained according to the spectral reflectance for each of the one or more types of feed; (b) estimating the amounts of the plurality of components for each of the one or more types of feed using a component estimation model that receives the spectral reflectance information obtained from the light reception results of at least some of the plurality of pixels as an input and outputs the amounts of the plurality of components; (c) outputting the component amounts of the plurality of components; A computer program that causes the processor to execute the above.

10. A computer program that causes a processor to execute a process for determining the blending amounts of multiple types of feed for multiple reared individuals, (a) using a spectroscopic camera including a spectral filter and an image sensor and capable of receiving light at a plurality of pixels arranged two-dimensionally, to capture images of the plurality of types of feed without crushing the feed for each type, thereby obtaining spectral reflectance from the light reception results at each of the plurality of pixels, and determining spectral reflectance information obtained according to the spectral reflectance for each of the plurality of types of feed; (b) estimating the amounts of the components of each of the plurality of types of feed using a component estimation model that receives the spectral reflectance information obtained from the light reception results of at least some of the plurality of pixels as an input and outputs the amounts of the components of each of the plurality of types of feed; (c) acquiring individual breeding information regarding the breeding status of each of the plurality of breeding individuals; (d) determining the required component amounts of the plurality of components for each of the plurality of reared individuals using a required component estimation model that receives the individual rearing information as input and outputs the required component amounts of the plurality of components; (e) determining the compounding amounts of the plurality of types of feed for each of the plurality of reared individuals from the component amounts of the plurality of components for each of the plurality of types of feed estimated in the process (b) and the required component amounts of the plurality of components for each of the plurality of reared individuals determined in the process (d); A computer program that causes the processor to execute the above.

Images