Water absorption state prediction device

Abstract

To provide a water absorption state prediction device capable of grasping a finish time of a water absorption step.SOLUTION: A water absorption state prediction device of the present invention comprises: an imaging unit that images grains immersed in water to obtain a captured image in response to a limited water absorption step in which grains absorb water in production of alcoholic beverage; an index calculation unit that calculates an index of the grains included in the captured image; a prediction index calculation unit that predicts a prediction index by predicting the index at a time after a capturing time of the captured image (future time); and a water absorption rate estimation unit that estimates a water absorption rate of the grains at the future time based on the prediction index.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a water absorption state predicting device that predicts the state of water absorption into grains (rice, etc.) in a water absorption process into grains in the production of alcoholic beverages and the like. [Background technology]

[0002] In Japan, sake has been produced and sold traditionally. Using rice and rice koji as the main ingredients, various types of sake have been produced by combining traditional production methods with new innovations. In the past, sake was classified as special grade sake or first grade sake, but in recent years, sake is no longer classified in this way and is instead produced and sold under categories such as junmai sake or ginjo sake.

[0003] Sake is a traditional Japanese alcoholic beverage and has been enjoyed in Japan for a long time. In addition to being enjoyed at celebratory occasions and banquets, it is also often enjoyed in everyday life. In the past, there were few sake producers (breweries), and many sakes contained brewer's alcohol due to the shortage of supplies after the war. This resulted in the limited popularity of sake. Furthermore, sake was mass-produced industrially, making it difficult to create distinctive sake.

[0004] In the past few decades, distinctive sake has been brewed through a focus on ingredients, the elimination of brewer's alcohol, and innovative production methods. Sake brewing that highlights the characteristics of each brewery has become widespread, especially among small breweries, and sake that takes into account regional characteristics and also pairs well with food is now being produced and sold.

[0005] As a result of these efforts, sake with a variety of characteristics is now being produced in various regions, and it is now possible to enjoy sake from a variety of regions by visiting each region or through nationwide distribution.

[0006] Furthermore, in recent years, sake has begun to be exported overseas. This is because, coupled with the global boom in Japanese cuisine, sake is becoming more accepted by overseas consumers. As a result, it is expected that the popularity of sake will continue to grow and the variety of sake will increase.

[0007] This type of sake is produced through a manufacturing process that includes various steps, including fermenting rice with koji. The manufacturing process includes certain steps that are set in stone to produce a more delicious sake, and producers will give some of the steps their own unique characteristics and use various ingenuity to bring out the unique characteristics of their sake. They also use ingenuity in the selection of ingredients, koji mold, and yeast used, as well as temperature and time management, to produce sake with a unique flavor.

[0008] In addition to sake, shochu and other alcoholic beverages are also produced using grains. These shochu and other alcoholic beverages are also produced using a manufacturing process that includes various steps.

[0009] The production of sake and other alcoholic beverages requires a water absorption process in which the rice or grains used as raw materials are hydrated and allowed to absorb the water. This process requires delicate control so that the rice absorbs the specified amount of water, and is called the "limited water absorption process." The raw rice is washed and then soaked in water. This soaking allows the rice to absorb water. By controlling the soaking time, the water absorption rate (amount of water absorbed) of the rice can be controlled.

[0010] The rate of water absorption into rice (sakamai) in sake production must be controlled very precisely, as the rate of water absorption affects the quality of the finished sake. In this respect, controlling the rate of water absorption into the rice during the limited water absorption process is extremely important.

[0011] Based on the master brewer's visual and experiential judgment, when the rice (sake rice) reaches a target water absorption rate, the rice is removed from the soaking water (or the water is discarded). In other words, the limited water absorption process must end and the absorbed rice must be moved on to the next process. In sake breweries, the decision to end the limited water absorption process was made by the master brewer and other workers based on manual labor, experience, and intuition. Naturally, much of this depended on the ability and experience of the worker, leading to problems of inconsistency and the time it took for workers to become skilled.

[0012] For this reason, techniques have been proposed for estimating the water absorption rate in the limited water absorption step in alcoholic beverage production (see, for example, Patent Documents 1 and 2). [Prior art documents] [Patent documents]

[0013] [Patent Document 1] Patent Publication No. 2021-99250 [Patent Document 2] Japanese Patent Publication No. 2020-41841 Summary of the Invention [Problem to be solved by the invention]

[0014] Patent Document 1 discloses a water absorption state evaluation device that includes an image data acquisition unit 10 that acquires image data of a group of target rice grains that are immersed in water, an image processing unit 11 that can perform predetermined processing on the image data acquired by the image data acquisition unit 10, an index determination unit 12 that determines, as one of the indices, an area-weighted average luminance determined based on the image data of the target rice grains, and a water absorption state evaluation unit 13 that evaluates the water absorption state of the target rice grains based on the indices determined by the index determination unit 12, where the area-weighted average luminance is the average luminance of the area occupied by the target rice grains in the image data of the target rice grains.

[0015] Patent Document 1 discloses a technology for estimating and calculating the water absorption state from images of soaked rice, with the aim of measuring the water absorption state of soaked rice.

[0016] However, Patent Document 1 has the following problems.

[0017] In Patent Document 1, the water absorption state is estimated by removing rice from the water absorption device during the water absorption process and taking an image. Therefore, there is a problem that the water absorption state of rice during the water absorption process in the water absorption device cannot be estimated as it is. Because it is necessary to remove some rice grains during absorption from the device or equipment containing water and rice for the limited water absorption process and take an image, there is a problem that the water absorption rate cannot be estimated for the rice itself undergoing the limited water absorption process.

[0018] There is a problem in that the water absorption rate is only estimated from a sample or academic perspective.

[0019] Furthermore, because the method is based on the premise of capturing images of groups of rice grains that are separated from one another, if the rice grains cannot be properly separated from the background, the brightness of the transmitted light used to estimate the water absorption rate will be inaccurate. This can lead to an inaccurate estimation of the water absorption rate. In fact, Table 2 of Patent Document 1 calculates a negative water absorption rate, revealing the problem of low accuracy.

[0020] Furthermore, Patent Document 1 estimates the current water absorption rate of rice. In other words, it does not have the function to predict the future water absorption rate. However, in the limited water absorption process in actual alcoholic beverage production, it is necessary to end the soaking and terminate the limited water absorption process when the water absorption rate reaches a predetermined value. This is because rice that has absorbed water at an optimal water absorption rate can be passed on to the next process.

[0021] However, in Patent Document 1, the water absorption rate is estimated after a processing time has elapsed since the image of the rice grain was captured. This means that the water absorption rate at the time of the image capture is estimated after a time lag. In other words, by the time the water absorption rate can be determined, the rice has already absorbed more water. Even if the rice is removed from the water at the time the estimated water absorption rate is determined, the target water absorption rate may have been exceeded.

[0022] This is a problem that has a negative impact on the quality of alcoholic beverages produced. The moment that workers can determine the rice's water absorption rate is after that rate has been reached, and even if they remove the rice from the water at that point, it is too late. The rice will absorb too much water, which will ultimately lower the quality of the alcoholic beverage produced.

[0023] Patent Document 2 discloses a change rate estimation device that includes a photographing device 2 that photographs rice immersed in water, an image data acquisition unit 11 that acquires two-dimensional image data of the rice photographed by the photographing device 2, viewed from the thickness direction of the rice, a crack rate calculation unit 16 that calculates the vertical crack rate by referring to the two-dimensional image data acquired by the image data acquisition unit 11, and an estimation unit 19 that acquires the vertical crack rate from the crack rate calculation unit 16 and estimates the rate of change in the shape of the rice in the thickness direction based on the vertical crack rate and a predetermined correlation between the vertical crack rate of the rice and the rate of change in the shape in the thickness direction.

[0024] Patent Document 2 also has the same problem as Patent Document 1.

[0025] As described above, conventional water absorption rate estimation devices have had problems such as difficulty in measuring the water absorption rate using the water absorption device itself, poor accuracy in estimating the water absorption rate, and a time lag between measurement and understanding the results.

[0026] In view of these problems, the present invention aims to provide a water absorption state prediction device that can determine the end time of the water absorption process by estimating the water absorption rate in the future based on images of rice during the water absorption process in a water absorption device. [Means for solving the problem]

[0027] In view of the above problems, the water absorption state predicting device of the present invention is A water absorption state prediction device that predicts the water absorption state by taking an image of the water and grains in the container of the water absorption device, It is soaked in water in accordance with the limited water absorption process that allows the grain to absorb water in alcoholic beverage production. Inside the container an imaging unit that captures an image of the grain; an index calculation unit that calculates an index of a grain included in a captured image; a prediction index calculation unit that predicts an index at a time (future time) after the imaging time of the captured image to predict a prediction index; a water absorption rate estimation unit that estimates the water absorption rate of the grain in the future based on the prediction index, The water absorption rate estimated by the water absorption rate estimation unit indicates the water absorption rate at a future time, which is a time after the time when the image is captured by the imaging unit. And, The index includes brightness in the captured image of the grain, The index calculation unit a brightness normalization unit that normalizes the brightness of the grain in the captured image using a normalization function in which the brightness of the maximum brightness in the reference image is set to 100 and the brightness of the minimum brightness is set to 0; The index calculation unit calculates the normalized brightness in the captured image of the grain, The results are output to the prediction index calculation unit. The prediction index calculation unit calculates the index as follows:

number

[0028] The water absorption state prediction device of the present invention can predict the water absorption state based on taking images of water and rice (grains) in the water absorption device. This makes it possible to apply it to actual alcoholic beverage manufacturing sites, rather than from a sample or academic perspective like the conventional technology.

[0029] In addition, the brightness of the rice grains in the captured image is used as an index for predicting water absorption. Normalizing the brightness can improve the accuracy of water absorption predictions, eliminating problems such as calculating a negative water absorption rate.

[0030] Furthermore, the change in the index obtained from the captured image from the time of capture to a future time is predicted. Based on this predicted index for future time, the water absorption rate for a future time can be predicted from the captured image. This eliminates the time lag when the worker can grasp the water absorption rate. Therefore, the limited water absorption process can be terminated at a more appropriate timing. Alternatively, the appropriate timing for terminating the limited water absorption process can be grasped. [Brief explanation of the drawings]

[0031] [Figure 1] FIG. 1 is a schematic diagram showing a part of the sake production process. [Figure 2] 1 is a block diagram of a water absorption state predicting device according to a first embodiment of the present invention. [Figure 3] FIG. 2 is a schematic diagram of an imaging unit including a light irradiation unit. [Figure 4] 1 is a captured image of sake rice captured by an imaging unit of the present invention. [Figure 5] FIG. 2 is a schematic diagram showing the relationship between the water absorption state of sake rice and brightness in the first embodiment of the present invention. [Figure 6] 3 is a schematic diagram showing the temporal relationship between captured images and estimation of water absorption rate in the first embodiment of the present invention. FIG. [Figure 7] 1 is a block diagram of a water absorption state predicting device according to a first embodiment of the present invention. [Figure 8] 4 is a graph showing calculation of predicted luminance in the first embodiment of the present invention. [Figure 9] 10 is a graph showing the correlation between predicted brightness and water absorption rate according to the first embodiment of the present invention. [Figure 10] FIG. 10 is a block diagram of a water absorption state predicting device according to a second embodiment of the present invention. [Figure 11] 1 is a graph showing the estimated water absorption rate as a function of time. [Figure 12] FIG. 10 is a block diagram of a water absorption state predicting device according to a second embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0032] The water absorption state prediction device according to a first aspect of the present invention includes an imaging unit that captures an image of grains immersed in water in a limited water absorption process for causing grains to absorb water in alcoholic beverage production, and an index calculation unit that calculates an index of the grain included in the captured image; a prediction index calculation unit that predicts the index at a time (future time) after the imaging time of the captured image to predict a prediction index; a water absorption rate estimation unit that estimates the water absorption rate of the grain at the future time based on the prediction index, The water absorption rate estimated by the water absorption rate estimation unit indicates the water absorption rate at a future time that is a time after the time when the image is captured by the imaging unit.

[0033] This configuration allows the operator to view the water absorption rate at a time later than the time of image capture. The water absorption rate can be displayed at a timing that takes into account the processing time from image capture to calculation of the water absorption rate. This allows the water absorption process to be completed without delay from the appropriate timing.

[0034] The water absorption state predicting device according to a second aspect of the present invention further includes a light irradiating unit that irradiates light onto the grain from the imaging unit side, in addition to the first aspect of the present invention, The image capturing unit captures an image of the light reflected from the grains irradiated by the light irradiating unit.

[0035] This configuration allows the grain batches to be properly imaged, and the index related to the water absorption rate to be properly calculated.

[0036] In a water absorption state predicting device according to a third aspect of the present invention, in addition to the first or second aspect, the index includes a luminance of the grain in the captured image, The index calculation unit The image capturing device includes a brightness normalization unit that normalizes the brightness of the grain in the captured image using a normalization function that sets the brightness of the maximum brightness in the reference image to 100 and the brightness of the minimum brightness to 0.

[0037] This configuration improves the accuracy of luminance calculation from captured images. In addition, since luminance is calculated relative to multiple captured images, the luminance value for a specific image does not deviate significantly.

[0038] In the water absorption state prediction device according to the fourth aspect of the present invention, in addition to the third aspect, the index calculation unit calculates the brightness of the normalized grain in the captured image and outputs the result to the prediction index calculation unit.

[0039] This configuration allows for the calculation of a more appropriate predicted brightness based on the relative brightness, leading to a more accurate estimation of the water absorption rate.

[0040] In the water absorption state predicting device according to the fifth aspect of the present invention, in addition to the fourth aspect, the predictive index calculation unit calculates the index by the following equation (1):

number

[0041] This configuration allows future indicators to be calculated with high accuracy.

[0042] In the water absorption state predicting device according to a sixth aspect of the present invention, in addition to the fifth aspect, when the index is the brightness, the predictive index is a predicted brightness at a future time from the time of imaging.

[0043] With this configuration, a predicted brightness that is correlated with the water absorption rate can be calculated, thereby making it possible to predict the water absorption rate in the future with high accuracy.

[0044] A seventh aspect of the present invention provides a water absorption state predicting device according to the fifth aspect of the present invention, further characterized in that the imaging unit captures images of the grain a plurality of times at different times, the index calculation unit calculates the index for each of the plurality of images; The prediction index calculation unit calculates a prediction index for a future time based on the plurality of indexes.

[0045] This configuration allows for more accurate prediction indicators to be calculated.

[0046] In the water absorption state predicting device according to an eighth aspect of the present invention, in addition to the fifth aspect, the predictive index calculation unit calculates the predictive index as a function of the predictive index at the future time.

[0047] This configuration makes it easy to grasp the water absorption rate in the future.

[0048] In the water absorption state predicting device according to a ninth aspect of the present invention, in addition to the first aspect, the water absorption rate estimating unit stores a correspondence relationship between the prediction index and the water absorption rate of the grain, The water absorption rate at the future time is estimated based on the correspondence relationship.

[0049] This configuration allows for highly accurate estimation of the water absorption rate.

[0050] The water absorption state prediction device according to the 10th aspect of the present invention, in addition to the features of the first aspect, further comprises an end time determination unit that determines an end time for ending the limited water absorption process based on the estimation result of the water absorption rate estimation unit.

[0051] With this configuration, the timing to end the limited water absorption step is not missed.

[0052] In the water absorption state prediction device according to the eleventh aspect of the present invention, in addition to the first aspect, the future time of the prediction index predicted by the prediction index calculation unit corresponds to the processing time from the imaging time of the captured image to the completion of the estimation of the water absorption rate by the water absorption rate estimation unit.

[0053] This configuration allows the time lag between capturing an image and estimating the result to be absorbed and the estimated water absorption rate can be obtained, preventing a delay in the completion of the limited water absorption process.

[0054] In the water absorption state predicting device according to a twelfth aspect of the present invention, in addition to the first aspect, the alcoholic beverage is sake, and the grain is sake rice.

[0055] This configuration makes it suitable for use in sake production.

[0056] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0057] (Embodiment 1)

[0058] (Overview) The water absorption state prediction device of the present invention can be used in the process of absorbing water into grains for alcoholic beverages, other beverages, processed foods, etc. It can be used in various water absorption processes that require grains to absorb water up to a target water absorption rate. For example, it can be used in the process of absorbing water into grains that are used as raw materials for alcoholic beverages in the production of alcoholic beverages.

[0059] Here, we will explain the limited water absorption process, which is a process for brewing sake using sake rice as a raw material, in which the sake rice is allowed to absorb water up to a target water absorption rate, as an example. The water absorption state prediction device of the present invention is used in this limited water absorption process. In other words, we will explain the case where the grain is sake rice and the alcoholic beverage is sake as an example.

[0060] First, an overview of the sake brewing process will be explained using Figure 1. Figure 1 is a schematic diagram showing part of the sake brewing process. The sake rice is washed in the rice washing step. The washed sake rice is then soaked in water. During this soaking step, the sake rice is allowed to absorb water up to the target water absorption rate for sake brewing. The target water absorption rate is determined by the sake brewer and the type of sake.

[0061] Once the limited water absorption process (hereafter referred to as the "water absorption process") is complete, the absorbed sake rice is removed from the water and moves on to the steaming process. After absorbing the water, the sake rice is steamed. Next, koji and yeast are added and the mashing process begins, after which the mashing process moves on to fermentation.

[0062] After fermentation, sake is produced through various processes, and improvements at each stage can be made to change the characteristics of the resulting sake.

[0063] The water absorption state predicting device of the present invention is used in this limited water absorption process.

[0064] FIG. 2 is a block diagram of the water absorption state predicting device according to the first embodiment of the present invention.

[0065] The water absorption state predicting device 1 includes an imaging unit 2, an index calculation unit 3, a prediction index calculation unit 4, and a water absorption rate estimation unit 5.

[0066] The imaging unit 2 captures an image of the sake rice soaked in water during the limited water absorption process. At this time, the imaging unit 2 captures the image while the water and sake rice remain contained in the container used for the limited water absorption process. In other words, the imaging unit 2 captures the image of the sake rice while the water and sake rice remain contained in the container and soaked. Unlike conventional technology, there is no need to remove a sample of sake rice from the container during water absorption and capture the image.

[0067] The imaging unit 2 can capture images during the limited water absorption process, so the water absorption state prediction device 1 can be applied to actual sake brewing sites. The water absorption state prediction device 1 can be deployed in many sake brewing sites. In other words, it helps reduce the burden on sake brewers and maintain quality.

[0068] Here, it is also preferable that the imaging unit 2 is equipped with a light irradiation unit 21 as shown in Fig. 3 to image the sake rice during the water absorption process (image the grain during the water absorption process). Sake rice 100 and water 110 are contained in a container 200. This allows the sake rice to absorb water inside the container 200. Fig. 3 is a schematic diagram of an imaging unit equipped with a light irradiation unit.

[0069] The light irradiation unit 21 irradiates light onto the sake brewer's rice from the imaging unit 2 side. In FIG. 3, the light irradiation unit 21 is shown irradiating the sake brewer's rice with light. As a result, the sake brewer's rice 100 is irradiated with light, and the sake brewer's rice reflects the light. The imaging unit 2 captures an image of the reflected light from the sake brewer's rice 100 irradiated by the light irradiation unit 21.

[0070] Figure 4 is an image of sake brewer's rice captured by the imaging unit of the present invention. As shown in Figure 4, the image captures the reflected light from the sake brewer's rice 100 irradiated with light. Even if the sake brewer's rice 100 contained in the container 200 is overlapping, the image captures the reflected light, so that the image can be obtained with the rice grains grasped.

[0071] The index calculation unit 3 calculates an index of the sake rice (grain) contained in the captured image. Here, the index includes the brightness of the sake rice in the captured image. The captured image can be expressed by brightness and chromaticity. The index calculation unit 3 calculates the brightness of the sake rice in this captured image as an index.

[0072] Because the image is taken using reflected light, it is possible to calculate brightness, which changes depending on the brightness of the sake rice. At this time, as shown in Figure 5, before the sake rice has absorbed water, there is little reflected light from the sake rice and the image is dark. In other words, brightness is low. As water absorption progresses, the sake rice contains more water and the reflected light increases, making the image brighter. In other words, brightness increases. Brightness, which is an index, has a correlation with the rate of water absorption.

[0073] Thus, brightness is suitable as an index showing the water absorption state of sake rice. Fig. 5 is a schematic diagram showing the relationship between the water absorption state of sake rice and brightness in embodiment 1 of the present invention. The index calculation unit 3 can calculate the brightness of sake rice as an index based on the captured image. This brightness as an index is the brightness at the time (timing) when the sake rice is imaged.

[0074] That is, it is an index at a timing earlier than the timing at which the water absorption state predicting device 1 produces a final result. For this reason, if the brightness calculated by the index calculating unit 3 is used as is to estimate the water absorption rate, the water absorption rate will be that at a time earlier than the timing at which the estimated result is obtained. For this reason, the index calculated by the index calculating unit 3 is used by the predictive index predicting unit 4, which will be described next.

[0075] The predicted index calculation unit 4 calculates a predicted index by predicting an index for a time ahead of the imaging time of the captured image (hereinafter referred to as "future time") using the index calculated by the index calculation unit 3. That is, the predicted brightness is calculated by predicting the brightness for a future time using the brightness calculated from the captured image.

[0076] The prediction index calculation unit 4 outputs the prediction index to the water absorption rate estimation unit 5. The water absorption rate estimation unit 5 estimates the water absorption rate of the sake rice based on the prediction index. The predicted brightness, which is the prediction index, has a correlation with the water absorption rate. This correlation can be defined using a function, a correspondence table, etc. Based on these definitions, the water absorption rate can be calculated and estimated from the predicted brightness.

[0077] This estimated water absorption rate indicates the water absorption rate at a future time, which is a time after the time when the image is captured by the imaging unit. Because this water absorption rate takes into account the time lag from the time of imaging until the water absorption rate is estimated, it is possible to accurately measure the timing when the target water absorption rate is reached and the sake brewer's rice should be removed from the container 200.

[0078] In the case of conventional technology, the past water absorption rate at the time of image capture was simply estimated, so if the estimated water absorption rate was the target water absorption rate, it was too late to remove the sake rice from the container.

[0079] In contrast, the water absorption state estimating device 1 of the present invention estimates the water absorption rate in the future from the image capture time. If the estimated water absorption rate is the target water absorption rate, it is not too late to remove the sake brewer's rice from the container at that point.

[0080] In addition, since the time of the image capture is recorded, the estimated time of the water absorption rate can also be determined. This allows the sake brewer to understand the state of the water absorption rate as an absolute time. Therefore, it is possible to know the time when the target water absorption rate is reached, and it is only necessary to remove the sake rice from the container at that time.

[0081] The imaging unit 2 can also capture images continuously over time. This allows images to be captured one after another over time, and the water absorption rates are estimated based on these captured images over time. As described above, each of these water absorption rates is the water absorption rate at a time in the future from the time of capturing the image.

[0082] By checking the successively outputted water absorption rates, the manufacturer can determine the timing to end the limited water absorption process (the timing to remove the sake rice from the container). For example, the time difference between the image capture time and the prediction index calculated by the prediction index calculation unit 4 corresponds to the processing time from the image capture time to the estimation of the water absorption rate. In this case, the sake rice can be removed from the container when the successively obtained water absorption rates reach the target water absorption rate. This allows the water absorption process to be completed without delay.

[0083] Of course, by calculating a future time prediction indicator with a time difference that takes into account the time it takes to remove the sake rice from the container, it is possible to end the water absorption process at a more accurate timing.

[0084] As described above, the water absorption state prediction device 1 in the first embodiment can grasp the target water absorption rate without delay in the limited water absorption process of sake rice in sake brewing. The water absorption rate can be grasped at a time that is understandable by the worker, with no time lag from the time of image capture. This makes it possible to improve and maintain the quality in the limited water absorption process in sake brewing. In particular, it can also reduce variations due to the proficiency of workers such as master brewers.

[0085] By achieving uniformity, improvement, and maintenance of quality in the limited water absorption process, it is possible to improve and maintain the quality of the sake produced. Of course, although sake production is used as an example here, the same applies when applied to other alcoholic beverages and foods.

[0086] Next, details and variations of each part will be explained.

[0087] (imaging unit) The imaging unit 2 captures an image while irradiating the sake brewer's rice 100 inside the container 200 with light from the light irradiating unit 21. This results in capturing an image of the light reflected by the sake brewer's rice 200. The more water the sake brewer's rice 200 absorbs, the brighter the reflected light becomes.

[0088] The imaging unit 2 may take images of the sake rice continuously over time to generate the captured images. The water absorption rate finally estimated by the water absorption state prediction device 1 is based on these captured images. Therefore, the water absorption rate is continuously estimated along the time axis. When multiple captured images are obtained along the time axis, estimated values ​​of the water absorption rate at estimated times in the future ahead of the image capture time are continuously obtained along the time axis.

[0089] This is as shown in Fig. 6. Fig. 6 is a schematic diagram showing the temporal relationship between captured images and estimation of water absorption rate in the first embodiment of the present invention.

[0090] Captured image 1 is obtained at imaging time 1. Finally, the water absorption rate estimation unit 5 calculates the water absorption rate at estimated time 1, which is in the future from imaging time 1, based on this captured image 1. The time difference between this imaging time 1 and estimated time 1 is the amount of time corresponding to the processing time as shown in FIG. 6. Water absorption rate 1 is calculated (estimated) at estimated time 1, which is a future time from imaging time 1 with the difference in the amount of time.

[0091] Therefore, the time when the worker grasps the water absorption rate 1 and the actual water absorption rate at that time are close to each other. If this water absorption rate 1 is the target water absorption rate, the worker will end the water absorption process at this estimated time 1.

[0092] 6, when captured images are obtained continuously along the time axis, the water absorption rates are also obtained continuously. Water absorption rate 2 is estimated at estimation time 2 corresponding to image capture time 2, water absorption rate 3 is estimated at estimation time 3 corresponding to image capture time 3, and water absorption rate 4 is estimated at estimation time 4 corresponding to image capture time 4.

[0093] The operator knows the target water absorption rate. Since the water absorption rate is continuously estimated along the time axis, when the operator finds a water absorption rate that is close to the target water absorption rate, the limited water absorption process can be terminated at that point.

[0094] For example, if the target water absorption rate is 70%, and water absorption rate 1 is 66.5%, water absorption rate 2 is 68%, and water absorption rate 3 is 69.9%, then at estimated time 3 for water absorption rate 3, the worker will end the limited water absorption process (remove the sake rice from the container or discard the water from the container). This is because there is a possibility that water absorption will have progressed too far by the time of the next water absorption rate 4.

[0095] In this way, by capturing consecutive images along the time axis, it is possible to end the water absorption process at the time when the water absorption rate is estimated to be close to the target water absorption rate, eliminating the time lag seen in conventional technology and realizing accurate timing.

[0096] Note that, here, the captured image 1 at the imaging time 1 is described as being based on a single image and its brightness, but in order to calculate the predicted brightness, which is a prediction index, the captured image 1 at the imaging time 1 may also be a group of multiple captured images 1 taken at slightly different times.

[0097] (Indicator calculation department) The index calculation unit 3 calculates an index for estimating the water absorption rate of the sake rice from the captured image. Here, the index includes brightness, which is related to the water absorption rate. It is appropriate to adopt brightness as the index. For this reason, the index calculation unit 3 calculates the brightness of the sake rice in the captured image.

[0098] Here, it is preferable that the index calculation unit 3 further includes a brightness normalization unit. Fig. 7 is a block diagram of the water absorption state prediction device according to the first embodiment of the present invention. The index calculation unit 3 includes a brightness normalization unit 31.

[0099] The brightness normalization unit 31 normalizes the brightness of the sake rice in the captured image using a normalization function that sets the brightness of the maximum brightness in the reference image to 100 and the brightness of the minimum brightness to 0. By normalizing, a brightness based on a relatively more appropriate difference is calculated. Using this brightness as a reference, the final water absorption rate is calculated more accurately.

[0100] The index calculation unit 3 outputs the brightness normalized by the brightness normalization unit 31 as an index to the prediction index calculation unit 4.

[0101] (Prediction index calculation part) The predictive index calculation unit 4 can calculate a future index as a predictive index by applying the received index to the function (logistic function) of (Equation 1). If the index is brightness, the predicted brightness for the future is calculated based on the brightness (which may be normalized brightness) calculated from the captured image itself. The predicted brightness is the brightness from the image capture time onward.

[0102]

number

[0103] The "y" on the left side of equation 1 is the predicted brightness in the future. The brightness calculated from the first captured image is input into the logistic function of equation 1 as "ymin," which is the minimum whiteness (minimum brightness, initial value). From the changes in brightness calculated from subsequent consecutive captured images, the least squares method is used to find "ymax," which is the maximum value (maximum brightness), and "k," which is the rate of increase. These ymax and k are input into the logistic function of equation 1.

[0104] Additionally, the amount of time from the start time of the limited water absorption process to the time of the prediction index you want to estimate in the future is entered as "t" in Equation 1. By entering this amount of time as "t", all the variables required for Equation 1 are entered.

[0105] As a result, the predicted brightness in the future can be calculated using Equation 1.

[0106] Furthermore, the imaging unit 2 captures multiple images at different times. The index calculation unit 3 calculates the brightness of each of these multiple images. The prediction index calculation unit 4 calculates a predicted brightness based on these multiple brightnesses. As shown in Equation 1, the predicted brightness for the future can be calculated based on the brightness of temporally consecutive captured images.

[0107] The prediction index calculation unit 4 calculates the predicted brightness as shown in Fig. 8. Fig. 8 is a graph showing the calculation of predicted brightness in the first embodiment of the present invention. As shown in the graph of Fig. 8, once a certain amount of actually measured brightness data has been accumulated, predicted brightness for the future time is calculated based on Equation 1. As shown in the graph, the data up to the solid line is actually measured brightness data, and the predicted brightness is shown by the dotted line thereafter.

[0108] In this way, the predicted brightness in the future is calculated, and the water absorption rate in the future can be estimated from the predicted brightness in the future.

[0109] In this way, when the index is brightness, the predicted brightness at a future time is calculated in the predictive index calculation unit 4. The water absorption rate estimation unit 5 estimates the water absorption rate at a future time based on the predicted brightness.

[0110] (Water absorption rate estimation section) The water absorption rate estimation unit 5 can estimate the water absorption rate from the predicted brightness. There is a correlation between brightness and water absorption rate. This correlation can be expressed by a function, a correspondence table, or the like. By inputting the predicted brightness into such a function or correspondence table, the water absorption rate estimation unit 5 can estimate the water absorption rate in the future.

[0111] The water absorption rate estimation unit 5 stores the correspondence (correlation) between the predicted brightness (prediction index) and the water absorption rate. This may be a function such as the one described above or a correspondence table. Based on this stored correspondence, the water absorption rate in the future can be estimated.

[0112] The water absorption rate in the future, taking into account the time when such processing will be performed, can be presented to the worker.

[0113] Fig. 9 is a graph showing the correlation between predicted brightness and water absorption rate in the first embodiment of the present invention. As shown in the graph of Fig. 9, by associating the predicted brightness with the correlation function between the predicted brightness and the water absorption rate, the water absorption rate can be estimated (calculated) from the predicted brightness.

[0114] As shown in Figure 8, the predicted brightness for the future is calculated from the accumulated brightness of multiple captured images. This predicted brightness is the brightness that would be obtained if the image were captured at a time in the future. The predicted brightness for the future can be used to estimate the water absorption rate for the future.

[0115] The water absorption rate is estimated using the correspondence relationship between the calculated predicted brightness and the corresponding water absorption rate.

[0116] The values ​​of ymin, ymax, k and the calculated predicted brightness used in Equation 1 are standardized. These standardized variables are multiplied by the rotation matrix and loading amount for the hyperplane. The value calculated by this multiplication is then subjected to an inverse transformation of the standardization using the mean and standard deviation of the water absorption recorded in advance. The value calculated by the inverse transformation becomes the estimated water absorption.

[0117] The water absorption rate recorded in advance here is the average value obtained from the analysis results of separately conducted experimental data. By collecting and analyzing the experimental data in advance, the average water absorption rate and other values ​​are calculated and stored. These values ​​are used to perform the above calculations and estimate the water absorption rate in the future.

[0118] As described above, the water absorption state predicting device 1 in the first embodiment can estimate the water absorption rate in the future. This allows the time lag between the time of image capture and the time the estimated result is obtained to be absorbed and the estimated result to be notified to the worker. As a result, the limited water absorption process can be ended at the timing when the target water absorption rate is determined, by removing the sake rice from the container or discarding the water from the container.

[0119] (Embodiment 2)

[0120] Next, a description will be given of a second embodiment. In the second embodiment, various variations will be described.

[0121] (Notification Department) Fig. 10 is a block diagram of a water absorption state predicting device according to the second embodiment of the present invention. The water absorption state predicting device 1 in Fig. 10 includes a notification unit 6. The notification unit notifies the estimation result of the water absorption rate estimation unit 5. By receiving this notification, the worker can understand the water absorption rate.

[0122] The notification unit 6 notifies by image, sound, text, email, visual display, etc. By receiving notification by these means, the worker can grasp the water absorption rate.

[0123] The notification unit 6 may not only notify the user of the water absorption rate, but also notify the user that the water absorption rate has approached the target water absorption rate. For example, this notification may be made by means of an image or sound. This allows the user to end the limited water absorption process at the appropriate time.

[0124] (Calculation by function) The prediction index calculation unit 4 calculates the predicted brightness as a function as shown in Fig. 8. The brightness at a future time is calculated as a function (calculated using a curve defined by a function, not a water absorption value). The predicted brightness may be calculated as a discrete value or may be calculated as a function.

[0125] The water absorption rate is estimated based on the predicted brightness (in the water absorption rate estimation unit 5). Therefore, the prediction index calculation unit 4 can calculate the water absorption rate as a function corresponding to the time axis. Of course, if a value at a certain time is extracted from the function, the water absorption rate can be obtained as a discrete value.

[0126] As time passes, the number of captured images increases, and the predicted brightness calculated as a function and the water absorption rate estimated from this are also obtained one after another.

[0127] If the water absorption rate is obtained by such a function, the operator can obtain the estimated water absorption rate in the future. As explained in the first embodiment, the water absorption process can be terminated when the target water absorption rate is obtained.

[0128] Alternatively, if the water absorption rate in the future is obtained as a function, the water absorption rate can be obtained as a function corresponding to the actual time. Figure 11 is a graph showing the estimated water absorption rate obtained as a function corresponding to this time axis.

[0129] The water absorption rate beyond the current time for the operator (a time ahead of the image capture time, which indicates the water absorption rate estimated in the future) is also estimated. Here, the target water absorption rate is determined in advance. As shown in FIG. 11, the time at which the target water absorption rate is reached can also be predicted from the function of the estimated water absorption rate. In FIG. 11, this is the time ΔT after the current time.

[0130] The time after ΔT can be grasped by the operator as an estimate of the time when the target water absorption rate will be reached. Since there is a limit to the future time that the prediction index calculation unit 4 can predict, there is also a limit to the future time that the water absorption rate can be estimated. However, as the water absorption process progresses, time and water absorption progress toward the right side of the graph in Figure 11.

[0131] That is, it becomes possible to predict the future time when the target water absorption rate will be reached. The right side of the graph in Figure 11 is obtained (the curve (straight line) of the graph gradually extends to the right of the time axis).

[0132] If the time when the target water absorption rate is reached can be determined, the worker can end the water absorption process at that time. Rather than checking the water absorption rate at each time and ending the water absorption process at that moment, there is an advantage in that the worker can predict the time when the water absorption process should end and complete the work with ample time to spare.

[0133] In this way, by being able to estimate the time when the target water absorption rate will be reached, the limited water absorption step can be ended at an appropriate timing.

[0134] It is also preferable to further provide an end time determination unit that can determine such an end time. Fig. 12 is a block diagram of a water absorption state prediction device in embodiment 2 of the present invention. The water absorption state prediction device 1 in Fig. 12 includes an end time determination unit 7. The end time determination unit 7 can determine the time at which the target water absorption rate is reached based on the graph in Fig. 11. In other words, it can determine the time at which the limited water absorption step should be ended.

[0135] The end time determination unit 7 determines the time when the target water absorption rate is reached as the end time, and notifies the operator via the notification unit 6. The operator ends the limited water absorption process at the end time determined by the end time determination unit 7.

[0136] Alternatively, if the termination of the limited water absorption step is automated, the result from the termination time determination unit 7 is notified to this automation process, which then terminates the limited water absorption step. This prevents failure to terminate the limited water absorption step.

[0137] The change in water absorption rate (the time required for the change) can also be calculated. This is calculated from the time change in estimated water absorption rate in the future obtained from the predicted brightness. This change in water absorption rate can be considered as the water absorption rate.

[0138] Being able to grasp this water absorption rate makes it easier for workers to carry out the limited water absorption process and leads to improvements. The correlation between the visual changes in sake rice and the actual water absorption rate can be grasped. This allows the correlation between the worker's visual observation and the actual water absorption rate to be matched with the data, making it easier for workers to understand the water absorption process. As a result, workers can improve their ability to grasp the water absorption state while they are working.

[0139] This will improve the skill level of the workers. In particular, there is an advantage in that the skills that could only be passed on orally or through experience from an expert can now be mastered by the workers by visually checking and understanding the water absorption rate through data.

[0140] Furthermore, a part or the whole of the water absorption state predicting device 1 described in the first and second embodiments may be realized by various devices such as a personal computer, a notebook computer, a tablet terminal, a smartphone, a mobile terminal, etc. Of course, it may also be realized by a dedicated device. It is also preferable that these devices are easily portable by the operator.

[0141] The water absorption state predicting device described in the first and second embodiments is an example for explaining the gist of the present invention, and includes modifications and alterations within the scope of the gist of the present invention. [Explanation of symbols]

[0142] 1. Water absorption prediction device 2. Imaging unit 21 Light irradiation unit 3 Indicator calculation part 31 Index normalization section 4. Prediction index calculation section 5 Water absorption rate estimation part 6 Notification section

Claims

1. A water absorption state prediction device that predicts the water absorption state by taking an image of the water and grains in the container of the water absorption device, an imaging unit that captures an image of the grains inside the container that are immersed in water in a limited water absorption process for absorbing water into the grains in alcoholic beverage production; an index calculation unit that calculates an index of the grain included in the captured image; a prediction index calculation unit that predicts the index at a time (future time) after the imaging time of the captured image to predict a prediction index; a water absorption rate estimation unit that estimates the water absorption rate of the grain at the future time based on the prediction index, the water absorption rate estimated by the water absorption rate estimation unit indicates a water absorption rate at a future time that is a time later than the time when the image is captured by the imaging unit, the index includes a brightness of the grain in the captured image, The index calculation unit a brightness normalization unit that normalizes the brightness of the grain in the captured image using a normalization function in which the brightness of the maximum brightness in a reference image is set to 100 and the brightness of the minimum brightness is set to 0; the index calculation unit calculates the normalized brightness of the grain in the captured image; outputting the results to the prediction index calculation unit; The prediction index calculation unit calculates the index as follows: [Equation 1] The prediction index can be calculated by applying the function A water absorption state prediction device, wherein the future time of the prediction index predicted by the prediction index calculation unit corresponds to the processing time from the image capture time of the captured image to the completion of estimation of the water absorption rate by the water absorption rate estimation unit.

2. a light irradiation unit that irradiates the grain with light from the imaging unit side; The water absorption state predicting device according to claim 1 , wherein the image capturing unit captures an image of light reflected from the grains irradiated by the light irradiating unit.

3. 3. The water absorption state predicting device according to claim 1, wherein when the index is the brightness, the predictive index is a predicted brightness at a future time from the time of image capture.

4. the imaging unit captures images of the grain a plurality of times at different times, the index calculation unit calculates the index for each of the plurality of images; The water absorption state predicting device according to claim 1 , wherein the predictive index calculation unit calculates a predictive index for a future time based on a plurality of the indices.

5. The water absorption state predicting device according to claim 1 , wherein the predictive index calculation unit calculates the predictive index as a function of the predictive index at the future time.

6. the water absorption rate estimation unit stores a correspondence relationship between the prediction index and the water absorption rate of the grain; The water absorption state predicting device according to claim 1 , wherein the water absorption rate at the future time is estimated based on the correspondence relationship.

7. The water absorption state predicting device according to claim 1 , further comprising an end time determining unit that determines an end time for ending the limited water absorption process based on the estimation result of the water absorption rate estimating unit.

8. 2. The water absorption state predicting device according to claim 1, wherein the alcoholic beverage is sake and the grain is sake rice.

Images