JP2019187259A - Culture support method, culture support program, culture support device and culture support system - Google Patents

Abstract

To provide useful information useful for culture by evaluating properly, a growth state of a plant such as a fruit tree.SOLUTION: A culture support method executed by a culture support system 1 uses a growth curve approximating to a growth state of fruit for evaluating a growth state of fruit which is being cultured, then, on the basis of the evaluation result, provides advice related to fruit being cultured. Therefore, the method can perform a useful culture support service by properly grasping a growth state of the fruit which is directly coupled to a profit of a farmer. The growth curve determines a size of fruit to an objective variable, and determines an integrated value of a leaf area to an explanatory variable. Then, the growth curve is determined on the basis of the size of the fruit and integrated value of the leaf area in fruit culture performed by an enthusiastic farmer. Therefore, a universal growth curve which does not depend on an environmental condition such as weather, can be obtained for performing culture support for approximating to a growth state of fruit similar to that of by the enthusiastic farmer.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a cultivation support method, a cultivation support program, a cultivation support apparatus, and a cultivation support system for evaluating the growth state of a crop such as a fruit tree and providing useful information based on the evaluation result.

  When cultivating crops such as fruit trees, it is important to appropriately grasp the growth state of the crops and to carry out appropriate work at an appropriate timing. It is also useful if the yield and quality can be accurately predicted. Therefore, it is useful if the growth state is appropriately evaluated during cultivation and necessary information can be provided appropriately based on the growth state. For example, when cultivating fruit trees such as apple trees, it is useful to periodically evaluate the growth state of the fruit and provide appropriate information on the fruit picking method, harvest time, and harvest size. Moreover, it is required to realize these at relatively low cost without labor.

  Patent Document 1 describes an information providing system that registers information useful for cultivation of fruit trees and garden trees in a database and provides users with information necessary for the growth of fruit trees and garden trees via the Internet. However, this information provision system only provides information on general stages of growth (seeding, tree planting, etc.) and information on diseases and fertilization. Therefore, with this system alone, it is impossible to grasp the growth state of the fruit tree cultivated by the user, and appropriate information cannot be provided.

  Patent Document 2 describes that an image of infrared light or near-infrared light is analyzed, and an index such as NDVI (normalized vegetation index) is calculated to predict the crop yield. In addition to NDVI, LAI (leaf area index) and GRVI (Green-Red Vegetation Index) have been proposed as indices indicating the growth state of plants. However, NVDI requires special equipment for measurement, and LAI has a heavy burden because measurement is a manual operation. On the other hand, GRVI can be obtained from an image by visible light that can be taken by a general-purpose camera, but the sky and the leaf may not be identified, and the leaf region cannot be extracted appropriately.

JP 2001-318963 A JP 2010-166851 A

As a new index for accurately identifying the sky and leaves, the present inventors have used GBVI (Green-Blue).
Vegetation Index) was proposed and a patent application was filed on March 24, 2017 (Japanese Patent Application No. 2017-58688). This application describes obtaining GBVI from an image of an apple tree and obtaining a growth curve indicating changes in the area of the leaf region and fractal dimension. The growth curve that approximates the change in the leaf area (leaf area) reflects the rise and fall of the leaves of the apple tree, and reflects the growth state of the apple tree.

In the above-mentioned application, it is proposed to grasp the growth status of apple trees using indices such as the leaf area obtained from GBVI and the fractal dimension of the leaf region. No direct evaluation is proposed. The growth curve of the above application is the evaluation of how the apple tree grows with the passage of days, with the horizontal axis representing the number of days elapsed and the vertical axis representing the fractal dimension of the leaf area or leaf region. is there. This growth curve changes every year according to the weather conditions because the weather conditions differ every year. Therefore, although the growth state of the apple tree in the current year can be evaluated from the shape of the growth curve, even if it is known that the growth in the current year is fast, etc., it is not clear whether the weather is the cause or the cultivation method The cultivation policy could not be obtained. Therefore, useful information on the cultivation method could not be provided.

  Information useful for fruit tree cultivation is information useful for harvesting a maximum number of fruits of a target size, for example. In the case of apples, the price of fruit is determined by size, color, etc., but size is particularly important. Generally, the target size for harvesting by farmers is the size that can be sold at the highest price. Basically, the larger one is the higher price. However, since the upper limit size that can be grown is determined depending on the variety, the farmer aims to grow as many fruits of that size as possible, aiming at a size close to the upper limit size. Since the size of the fruit is mainly increased by the nutrients produced by the photosynthesis of the leaves, management of the number of leaves and fruits is important throughout the year. For example, when the apple cultivar is “Fuji”, the farmer picks leaves so that one fruit is left in approximately 4 top buds (about 10 leaves per top bud). And as a result, harvest the maximum number of fruits of the target size.

  Conventionally, fruit tree cultivation guidelines are based on tacit knowledge such as cans and experiences of serious farmers, and there are no quantitative work guidelines. In particular, the weather conditions vary from year to year, so even if the time (month and day) is the same, it is natural that the degree of fruit growth changes from year to year. No guidance has been proposed. In addition, as described above, Japanese Patent Application No. 2017-58688 of the present inventors also proposes a universal growth curve that does not vary as long as the same work as that of a serious farmer is performed regardless of environmental conditions such as weather. Not done. Therefore, information useful for fruit tree cultivation could not be provided.

  In view of the above problems, an object of the present invention is to provide information useful for cultivation by appropriately evaluating the growth state of fruit trees.

  Another object of the present invention is to provide information useful for cultivation by appropriately evaluating the growth state of crops other than fruit trees.

  In order to solve the above-mentioned problems, the cultivation support method of the present invention evaluates the growth state of the portion to be harvested in the crop under cultivation using a growth curve that approximates the growth state of the portion to be harvested in the crop, and evaluates Based on the result, a cultivation support service for the crop under cultivation is performed.

  The growth curve of the present invention approximates the growth state of the portion to be harvested of the crop. Therefore, by using this growth curve, it is possible to appropriately grasp the current growth state of the portion directly related to the profit of the farmer regarding the crop being cultivated. Moreover, in this invention, based on the evaluation result using a growth curve, cultivation support services, such as provision of the advice regarding the crops during cultivation, are performed. If it does in this way, the cultivation support according to the present growth state can be performed rather than simply providing knowledge of a general cultivation method. Therefore, since useful cultivation support can be performed, the farmer can perform an appropriate operation at an appropriate timing based on the provided service. Therefore, the profit of the farmer can be increased.

In the present invention, the growth curve is determined using the size of the portion to be harvested as an objective variable, and an integrated value of a leaf area that is an area of a leaf region in which the crop is taken as an explanatory variable. Further, when the crop to be cultivated is a fruit tree and the portion to be harvested is a fruit, the growth curve uses the size of the fruit as a target variable, and the area of the leaf region (leaf) in the image of the crop The integrated value of (area) is determined as an explanatory variable. And the said growth curve is determined based on the data of the integrated value of the size of the said fruit, and the said leaf area in fruit tree cultivation which a serious farmer performed.

  In fruit tree cultivation, an index that contributes most to fruit growth is leaf area. The leaf area in the present invention is not an actual leaf area, but an area of a portion that is subjected to photosynthesis under light. For example, if the weather conditions are bad and sufficient sunlight cannot be obtained, the area of the portion that is exposed to sunlight and performing photosynthesis is also reduced. As a result, photosynthesis is not sufficiently performed, and fruit growth is also delayed. In the present invention, the growth curve with the integrated value of the leaf area as the explanatory variable and the fruit size as the objective variable shows the relationship between the amount of photosynthesis and the fruit growth, so it is universal regardless of environmental conditions such as weather. is there. In addition, the growth curve of the present invention is a curve when the balance between the leaf area and the number of fruits is optimally maintained by fruiting work by a serious farmer, so if a person performs inappropriate fruiting work, It changes when the balance between the area and the number of fruits collapses. That is, this growth curve is a universal growth curve that does not fluctuate as long as the same work as that of a serious farmer is performed.

  Therefore, this growth curve can be used for advice on cultivation policy. For example, it can be used for the evaluation of fruit picking work. In addition, as described above, this growth curve is universal regardless of environmental conditions such as weather, and thus can be used for fruit growth prediction, harvest time prediction, harvest size prediction, and the like. This growth curve shows the relationship between the amount of photosynthesis and fruit growth, so if this growth curve changes despite proper work, the cause of disease or pests Conceivable. Since the occurrence of diseases, pests and the like can be confirmed visually, if such a situation occurs, provision of useful information using this growth curve can be avoided. Thus, since useful cultivation support can be performed, the profit of a farmer can be increased.

  In the present invention, the data of the integrated value of the fruit size and the leaf area are collected over a plurality of years, and the growth curve is obtained by normalizing the average size of the fruits measured every year, and then normalizing the plurality of the plurality of normalized fruits. It is preferably a universal curve obtained from the size of the fruit in the year. When the size of each fruit is normalized, the influence of the apparent size of the fruit size, that is, the effect that the distance between the fruit to be measured and the camera is different for each fruit can be excluded. In addition, if you normalize the average fruit size measured every year, you can create a growth curve using the normalized fruit sizes of multiple years at the same time. A growth curve can be obtained. Therefore, even if environmental conditions, such as a weather, differ every year, useful cultivation support can be performed.

  In the present invention, the integrated value of the leaf area is a value obtained by integrating the maximum value of the leaf area obtained at different times of the day every day. By using the maximum value in one day, it is not affected by changes with time in the day, and changes in the amount of photosynthesis due to daily weather changes are reflected in the integrated value of the leaf area . Therefore, a growth curve representing the relationship between the amount of photosynthesis and fruit growth can be obtained.

  In the present invention, a logistic curve can be used as the growth curve. A logistic curve is a typical growth curve. The present inventors performed a curve approximation with a logistic curve using the fruit size as an objective variable and the integrated value of the leaf area as an explanatory variable. Other growth curves can also be used. For example, a Gompertz curve which is another growth curve can be used.

In the present invention, from the image of the fruit tree being cultivated, to determine the integrated value of the leaf area until the time of evaluating the growth state, based on the value on the growth curve corresponding to the integrated value, The growth state of the fruit in the fruit tree during cultivation can be evaluated. In this way, fruit growth prediction, harvest time prediction, harvest size prediction, and the like can be performed on the fruit tree being cultivated. In addition, useful information such as fruit picking method, harvest time, and harvest size can be obtained.

  In the present invention, the fruit size prediction at the time of harvest can be performed as the cultivation support service. Alternatively, as the cultivation support service, advice on the fruit picking operation can be performed. Moreover, the harvest date prediction of the fruit can be performed as the cultivation support service. By using a growth curve with the size of the fruit as the objective variable, it is possible to predict which size the fruit will become at what time. Therefore, the size at the time of harvest and the date of harvest can be predicted. In addition, based on the maximum size that can be harvested, the guidelines for fruit picking can be determined. For example, when the maximum size that can be harvested is smaller than the target size, advice can be given that the fruit should be picked.

  In the present invention, the size of the fruit in the fruit tree being cultivated is measured for a predetermined period, an approximate straight line that approximates the change in the size of the fruit is obtained, and the slope of the approximate straight line and the slope of the tangent line of the growth curve are determined. It is preferable to compare and give advice on additional fruit removal work based on the comparison result. In this way, the growth rate of the fruit in the growth curve (the slope of the tangent line of the growth curve) can be compared with the growth rate of the fruit in the plant that is actually being cultivated (the slope of the approximate line). Advice can be given considering the difference. For example, when the fruit growth rate in the plant actually cultivated is faster than the fruit growth rate in the growth curve, it can be advised that no additional fruit removal is necessary.

  In the present invention, it is preferable that the size of the fruit is measured by extracting the fruit from a color image obtained by photographing the fruit tree. If it does in this way, since it is not necessary to measure the size of a fruit on the spot, a burden can be eased.

  In the present invention, the size of the fruit is preferably measured using an AI that has learned the process of extracting the fruit using a color image obtained by photographing the fruit tree. The present inventors have confirmed that AI can learn a process of extracting fruits from a color image obtained by photographing a fruit tree, using a color image obtained by photographing the fruit tree as teacher data.

In the present invention, a single pixel or unit region of the color image using a G image that is an image in a green wavelength region of a color image obtained by photographing the plant and a B image that is an image in a blue wavelength region of the color image. For each, GBVI is calculated by the following formula (A),
GBVI = (GB) / (G + B) (A)
For each single pixel or unit region, generating a GBVI image by binarizing the pixel value calculated by the formula (A) based on a fixed threshold value, and obtaining the leaf area from the GBVI image preferable. When GBVI is used, the sky and the leaf can be distinguished, and the leaf region can be appropriately extracted. Therefore, the leaf area can be obtained with high accuracy. Further, since the threshold value is fixed without changing the threshold value for each image, the area of the portion where photosynthesis is performed can be obtained as the leaf area. Therefore, using the integrated value of the leaf area, the growth state of a portion to be harvested such as a plant or fruit can be approximated. GBVI can be obtained from an image taken with a general-purpose camera. Therefore, it is possible to grasp the growth state of a portion to be harvested such as a plant or a fruit using a simple and inexpensive device.

  Next, the cultivation support program of the present invention provides the computer with the cultivation support service based on the evaluation process of the growing state of the crop under cultivation based on the growth curve in the cultivation support method and the evaluation result. And processing for determining advice information to be performed.

  In the cultivation assistance program of this invention, it is preferable to make a computer perform the measurement process of the said fruit size in said cultivation assistance method. If the size of the fruit can be automatically measured, the burden on the user can be reduced.

  Next, the cultivation assistance apparatus of this invention has a memory | storage part which memorize | stores said cultivation assistance program, and a control part which reads and runs the said cultivation assistance program from the said memory | storage part, It is characterized by the above-mentioned.

  Next, the cultivation support system of the present invention includes a camera that captures a color image with visible light, a data storage unit that stores the color image captured by the camera, and a predetermined capturing schedule that controls the camera. The color image is captured by the data storage unit, the data measurement unit that stores the color image in the data storage unit, and the color image read from the data storage unit, and the color image read out from the data storage unit, 16. The integrated value of the leaf area in the cultivation support method is calculated, and the cultivation support is calculated using the integrated value of the leaf area and the growth curve in the cultivation support method according to any one of claims 2 to 15. And a data processing unit for determining advice information provided in the service.

  In this way, if a system is built that controls the camera to photograph plants according to a predetermined schedule and automatically accumulates color images in time series, the data processing unit accesses the data storage unit and stores it. The integrated value of the leaf area can be automatically obtained using the color image. In addition, the fruit size and growth rate can be automatically obtained using the accumulated color image. Therefore, the burden on the user can be reduced.

  In constructing such a cultivation support system, the data processing unit is provided in a server connected to a network, and the server provides the cultivation support service to a notification destination registered in advance via the network. It can be configured to perform the notification. In this way, the user can be encouraged to use the advice information.

  In this case, the server uses the color image read from the data storage unit based on a request from a terminal connected to the server via the network to monitor the growth state of the crop. Is preferably displayed on the terminal. Thus, by displaying the captured image itself on the terminal, it is possible to more accurately grasp the growing state of the crop.

  In the cultivation support system of the present invention, the server may include all or part of the color image and the advice information associated with each of a plurality of users, or all or one of the plurality of users. Managed as shared information that can be viewed by the department. If it does in this way, each user (farmer) can refer to the growth state and advice information in other users (farmers). Therefore, it is possible to provide a cultivation support system that is more useful for the user by utilizing the accumulated data.

  In this case, it is preferable that the server compares and displays all or part of the color image and the advice information associated with each of a plurality of users. If the comparison display can be performed, the growth state can be easily compared. In addition, it becomes easy to compare the results of work performed according to the advice information. In particular, when performing comparison display, statistical information such as averages and histograms is displayed for those that can be quantified, and distribution information such as frequency when divided into several categories is displayed for those that cannot be quantified. It is preferable. In this way, each user can grasp the state of many crops and the collective knowledge of many users, and can make decisions with reference to them.

  Further, it is preferable that the server manages the comment by receiving registration of a comment for the color image and the advice information. If a comment function is provided, information exchange and advice can be performed between users. Therefore, it is possible to provide a cultivation support system that is more useful for each user (farmer).

  Furthermore, it is preferable that the server receives registration of a work schedule from each of the plurality of users and manages the work schedule as shared information. In this way, the work schedule can be shared and adjusted between users. Therefore, it is possible to provide a cultivation support system that is useful for users to work in cooperation with each other.

It is a flowchart of the cultivation assistance method to which this invention is applied. It is a flowchart of the determination method of a growth curve. It is a structural example of the data measurement part installed in a plantation. It is a flowchart of the method of calculating | requiring the integrated value of a leaf area, and the fractal dimension of a leaf area | region. It is explanatory drawing which shows the production | generation method of a GBVI image. It is explanatory drawing which shows a time-dependent change of a color image, a GBVI image (grayscale display), a GBVI image (binary display), the area of a leaf area, and a fractal dimension. It is a graph which shows a time-dependent change of a leaf area and a fractal dimension. It is a graph which shows the time-dependent change of the size of a fruit. It is explanatory drawing of a universal growth curve (universal curve) It is explanatory drawing of the growth curve determined using four types of explanatory variables. It is explanatory drawing of the cultivation assistance service using a growth curve. It is explanatory drawing of the automatic detection of the fruit by AI, and the automatic calculation of an approximate straight line. It is explanatory drawing of the necessity determination of fruit extraction using an approximate straight line. It is a block diagram of a cultivation support system and a cultivation support system to which the present invention is applied. It is explanatory drawing of the Example of the cultivation assistance method by the cultivation assistance system to which this invention is applied. It is an example of the terminal using a cultivation support system.

  Embodiments of a cultivation support method, a cultivation support program, a cultivation support apparatus, and a cultivation support system to which the present invention is applied will be described below with reference to the drawings.

(Background and Summary of the Invention)
As described in the section of the problem to be solved by the invention, the present inventors have proposed a growth curve that approximates the growth state of fruit trees in Japanese Patent Application No. 2017-58688. In the growth curve of Japanese Patent Application No. 2017-58688, the horizontal axis is the number of days elapsed, the vertical axis is the area of the leaf region or the fractal dimension, and it was evaluated how the fruit tree grew with the passage of days. . This growth curve changes every year according to the weather conditions because the weather conditions differ every year. Therefore, the purpose was to evaluate the growth state of the fruit tree in the year.

On the other hand, the growth curve used in the present invention is not a approximation of the growth of the entire fruit tree, but a growth curve that approximates the growth state of the fruit that directly affects the profits of the farmer. As was done in Japanese Patent Application No. 2017-58688, the shape of the growth curve changes every year if the growth curve is obtained with the elapsed time on the horizontal axis and the fruit diameter actually measured on the vertical axis. Can be evaluated. For example, it can be evaluated that this year's fruits grow faster than the average year. However, even if such a growth curve was created and it was found that the growth of the fruit was fast, it was not possible to obtain the cultivation policy because it was not known whether the weather was the cause or the cultivation method. Therefore, the cultivation support method of the present invention is a method for obtaining useful information such as a fruit picking method, harvest time, harvest size, etc., using a universal fruit growth curve that does not vary as long as the same work as that of a serious farmer is performed. ing.

  Since the index that contributes most to the growth of the fruit is the area of the leaf region, the growth curve of the present invention has the horizontal axis as the integrated value of the area of the leaf region (leaf area) and the vertical axis as the fruit diameter. Here, the present invention is characterized in that the definition of the area of the leaf region (leaf area) is not the actual area of the leaf, but the area of the part that is subjected to photosynthesis under light. For example, if the weather conditions are bad and sufficient sunlight cannot be obtained, the area of the portion that is exposed to sunlight and performing photosynthesis is also reduced. As a result, photosynthesis is not sufficiently performed, and fruit growth is also delayed. Thus, since the growth curve of the present invention shows the relationship between the amount of photosynthesis and fruit growth, it is universal regardless of environmental conditions such as weather.

  Further, the growth curve of the present invention is a curve when the balance between the leaf area and the number of fruits is optimally maintained by fruiting work by a serious farmer. Therefore, if a human performs an inappropriate fruiting operation and the balance between the leaf area and the number of fruits is lost, the growth curve changes. Therefore, this growth curve can be used for the evaluation of the fruiting operation. In addition, as described above, the growth curve of the present invention is universal regardless of environmental conditions such as weather, and therefore can be used for fruit growth prediction, harvest time prediction, harvest size prediction, and the like.

  Moreover, the growth curve of the present invention represents the relationship between the amount of photosynthesis and fruit growth. Therefore, if this growth curve changes despite proper work, there may be a cause of disease or pest. Since the occurrence of diseases and pests can be visually confirmed, when such a situation occurs, useful information using this growth curve is not provided.

(Cultivation support method for fruit trees)
Hereinafter, with reference to FIGS. 1-16, embodiment of this invention which aimed at fruit trees (apple tree) is described. Fig.1 (a) is a flowchart of the cultivation support method of the fruit tree to which this invention is applied. As shown to Fig.1 (a), in the cultivation assistance method of this invention, the growth curve which approximated the growth state of the fruit in a fruit tree is determined previously (step ST1). Next, the growth state of a fruit is evaluated using a growth curve (step ST2). And based on the evaluation result of a growth state, the cultivation support service of a fruit is performed (step ST3). FIG. 1B is a flowchart when the present invention is extended to a cultivation support method for all agricultural products, and the description thereof will be described later.

  FIG. 2 is a flowchart of a method for determining a growth curve. In this embodiment, the integrated value of the leaf area is used as an explanatory variable for determining the growth curve of apple tree fruit. Moreover, the measured value of the size of a fruit is used as an objective variable. The leaf area is the area of the leaf area of the image taken from the apple tree, but since the threshold value for extracting the leaf area from the image is constant, photosynthesis is performed not by the actual leaf area but by light. It is the area of the part. Since fruits are mainly grown by nutrients produced by leaf photosynthesis, the integrated value of the leaf area can be an indicator of fruit growth. The leaf area (the area of the leaf region) is an index proposed by the present inventors in Japanese Patent Application No. 2017-58688, and the integrated value of the leaf area is an integrated value of the leaf area every day.

In this embodiment, the calculation of the leaf area and the measurement of the fruit size are performed using a color image obtained by photographing a fruit tree (apple tree) in the farm. Therefore, as shown in FIG. 2, in step ST11, a camera is installed in the farm, and a fruit tree is photographed to obtain a color image. Next, in step ST12, the leaf area is obtained from the color image taken in step ST11, and in step ST13, the integrated value of the leaf area is obtained. On the other hand, in step ST14, the size of the fruit is obtained using the color image taken in step ST11. The size of the fruit may be measured by manually extracting the fruit from the color image or by a program for extracting the fruit from the image.

  In this embodiment, the integrated value of the leaf area is obtained over the entire period in which the leaves and fruits grow in one year. In the shooting performed in 2016, the inventors performed the shooting of step ST11 every hour from 8 to 20 o'clock every day in principle during the period from April 1 to December 15. As a result, 3281 color images were obtained. In step ST12, for the shooting date during this period, first, the leaf area of all images shot on that day was calculated, and then the maximum leaf area of that day was determined as the leaf area of that day. In step ST13, the integrated value of the leaf area from the start date of the imaging period was calculated every day. In addition, the fruit size, which is the objective variable, was measured from images taken over the period from the time when the fruits could be recognized from the taken images to the fruit harvest. In the example of 2016, the present inventors have manually picked six fruits from images taken at 12:00 on each shooting day in the period from June 1 to the harvest date (November 21). Was extracted, and the unit was measured as a pixel. The six fruits measured here are fixed in the period from June 1 to the harvest date, and if the fruit is hidden in the leaves etc. in the image taken at 12:00, an image at another time of the day is displayed. Used to measure size.

  FIG. 3 is a configuration example of the data measuring unit 10 installed in the farm. The inventors of the present invention performed automatic photographing of a color image of a fruit tree using the data measuring unit 10 of FIG. As shown in FIG. 3, the data measurement unit 10 includes a camera 11, a measurement computer 12, a communication unit 13, and a power supply unit 14. Since the configuration of the data measuring unit 10 is the same as that described in Japanese Patent Application No. 2017-58688, detailed description thereof is omitted. That is, the camera 11 is a general-purpose single-lens reflex camera, and the settings are autofocus and aperture priority (F5.6). The camera 11 is positioned and installed so that the plant (fruit tree) to be evaluated is displayed in the viewfinder.

  The measurement computer 12 reads and executes a control program by a treatment device such as a CPU, thereby controlling the camera 11 to take an image of a plant on a predetermined schedule, and taking an image taken via the communication unit 13. It functions as a measurement unit that accumulates in the cloud server 20 (see FIG. 14A). In this embodiment, a microcontroller is used as the measurement computer 12. In this embodiment, an LTE modem is used as the communication unit 13.

  Here, before explaining the steps (steps ST15 to ST17) for determining the growth curve from the integrated value of the leaf area and the size of the fruit tree (steps ST15 to ST17), the integrated value of the leaf area, which is an index adopted as an explanatory variable in the present invention, The fractal dimension of the leaf region that was not considered but was considered as a candidate for the explanatory variable will be described.

(Leaf area and fractal dimension of leaf area)
FIG. 4 is a flowchart of a method for obtaining the integrated value of the leaf area and the fractal dimension of the leaf region. The fractal dimension of the leaf region is an index proposed in Japanese Patent Application No. 2017-58688 together with the leaf area. FIG. 5 is an explanatory diagram showing a method for generating a GBVI image. Since the method for obtaining the leaf area and the fractal dimension of the leaf region is the same as the method described in Japanese Patent Application No. 2017-58688, only the outline thereof will be described. Leaf area is determined from GBVI. GBVI (Green-Blue Vegetation Index) is obtained from a G image that is an image in the green wavelength region of a color image obtained by photographing a plant with visible light and a B image that is an image in the blue wavelength region of the color image.

As shown in FIG. 4, in order to obtain the leaf area, first, a color image of a plant is acquired (step ST21). Then, GBVI is calculated for each pixel (unit region) of the color image (step ST22). As shown in FIG. 5, the main area in which the apple tree to be evaluated is shown in the original image is cut out as an analysis target. For example, a color image of 2048 × 2048 pixels is cut out. The color image (RGB image) includes a G image that is an image in the green wavelength region, a B image that is an image in the blue wavelength region, and an R image that is an image in the red wavelength region. GBVI is calculated by the following formula (A) for each pixel, for example, using the luminance value of each pixel (each unit region) of the G image and B image.
GBVI = (GB) / (G + B) (A)

  GBVI obtained from the formula (A) is in the range of −1 to 1. In step ST23, the GBVI of each pixel (each unit region) is converted into a luminance value (pixel value) in the range of 0 to 255, thereby generating an 8-bit (256 gradation) grayscale image. In this specification, the gray scale image obtained in step ST23 is a GBVI image (gray scale display). Next, in step ST24, binarization processing is performed for converting the GBVI image (grayscale display) obtained in step ST23 into a monochrome binary image. In this specification, the monochrome binary image obtained in step ST24 is a GBVI image (binary display).

  Conversion (binarization processing) from a GBVI image (grayscale display) to a GBVI image (binary display) is performed using a predetermined threshold. In the present invention, the threshold value used for the binarization process is a fixed value, and all images are binarized using the same threshold value. For example, in the example of generating a GBVI image (binary display) using a color image of an apple tree, the value of the threshold Th is set to 150. As shown in FIG. 5, the white region of the GBVI image (binary display), that is, the pixel region on the high luminance side of the binarized image, can successfully extract the apple tree leaf region in the color image.

In step ST25, as shown by the following formula (B), the number of white pixels in the GBVI image (binary display) is calculated, and the ratio of the white pixels to the total number of pixels (2048 × 2048) is calculated. Thus, the area of the leaf region is obtained.
Leaf area (%) = number of white pixels / total number of pixels (B)
In step ST27, the fractal dimension of the leaf region is obtained. The fractal dimension of the leaf region is an index indicating the complexity of an image having self-similarity. The fractal dimension of the leaf region can be calculated using a box count method from an image of 2048 × 2048 pixels used when obtaining the leaf area. FIG. 6 is an explanatory diagram showing temporal changes in color image, GBVI image (grayscale display), GBVI image (binary display), leaf area, and fractal dimension. April 15th, May 15th , Images taken at 11:00 on June 15, July 15, August 16, September 15, October 15, November 16, and December 15 are shown. In this way, the leaf area and fractal dimension can be obtained for all photographed color images.

(Determining the representative value of leaf area and representative value of fractal dimension of leaf region)
FIG. 7 is a graph showing changes in leaf area and fractal dimension with time, and the horizontal axis represents the number of days elapsed from April 1st. It is data of 2016, and the typical value in one day was shown about each of a leaf area and a fractal dimension. When obtaining the integrated value from the fractal dimension data of the leaf area and the leaf area, the present inventors use the representative value of each photographing day instead of using the data of all the time zones (8:00 to 20:00). It was decided. The reason for this is that data fluctuations due to daily weather changes affect fruit growth, but changes with time always exist even without weather changes and do not affect fruit growth.

(Calculation of integrated value)
In this embodiment, the representative value of the fractal dimension of the leaf area and the leaf region is the maximum value of the day, but from the average value of the values calculated from all images taken on that day or from images taken within a certain time of the day It has been confirmed that the maximum, median, and minimum values of the calculated values can be substituted. In steps ST26 and ST28, the integrated value is obtained using the representative value (maximum value). That is, in step ST26, the daily representative value (maximum value) of the leaf area is integrated every day to obtain the integrated value of the leaf area. In step ST26, the daily representative value (maximum value) of the fractal dimension of the leaf area is integrated every day to obtain an integrated value of the fractal dimension of the leaf area.

(Determination method of growth curve)
With reference to FIG. 2, the latter half of the growth curve determination method will be described. As described above, in steps ST12 to ST13, the integrated value of the leaf area, which is an explanatory variable, is obtained over the entire period (for example, from April 1 to December 15) in which leaves and fruits grow within one year. Moreover, in step ST14, the size of the fruit which is an objective variable is calculated | required over the period (for example, June 1 to a harvest day) in which a fruit can be confirmed. In step ST15, curve approximation using the logistic curve equation shown in equation (C) is performed using these data. That is, a logistic curve that is a typical example of a growth curve of an organism is employed as a growth curve model.

  FIG. 8 is a graph showing the change in fruit size over time, where the horizontal axis represents the number of days elapsed from April 1 and the vertical axis represents the fruit size (apple radius). The size of the fruit is measured by manually extracting the fruit from an image of 2048 × 2048 pixels used for calculating the leaf area and setting the unit as a pixel, and measuring the radius. These are data in which a plurality of (in this embodiment, six) fruits are measured. 8A and 8B are data for 2016, and FIGS. 8C and 8D are data for 2017. FIGS. 8A and 8C show changes in individual fruit sizes in each year, and FIGS. 8B and 8D show changes in average fruit sizes in each year. As shown in FIG. 8, individual fruit sizes vary widely, and the average fruit size varies from year to year. This occurs because the fruit in the image to be measured is not constant from the camera. That is, the fruit in the image becomes larger as it is closer to the camera and smaller as it is farther, so the actual size is unknown. However, since the size of the fruit that can be measured from the image increases with the growth of the fruit, it is possible to grasp a relative change. In step ST14, the average fruit size in each year shown in FIGS. 8B and 8D is obtained.

  In step ST15, a growth curve is determined from data on the growth state of leaves and fruits for one year. The growth curve determined in step ST15 (the growth curve determined from single-year data) can be used for cultivation support, but fruit size is normalized in step ST16, and data for multiple years is used in step ST17. A universal growth curve (universal curve) can also be created.

(Normalization processing)
The fruit size shown in FIG. 8 is an apparent size depending on the distance from the camera of the fruit to be measured. The normalization process in step ST16 is a process for eliminating the influence of the distance from the fruit camera. In step ST16, among the coefficients a, b, and c of the growth curve determined every year, the coefficient a which is an asymptote of the logistic curve (that is, the upper limit of the fruit size) is used to determine the fruit size as the objective variable. Perform normalization. For example, by dividing the fruit size data for each year by the coefficient a for that year, the fruit sizes in 2016 and 2017 can all be normalized to be in the range of 0 to 1.

(Create universal curve)
FIG. 9 is an explanatory diagram of a universal growth curve (universal curve). FIG. 9A shows the result of normalizing the growth curve created using the 2016 data. FIG. 9B shows the result of normalizing the growth curve created using the 2017 data. And FIG.9 (c) is the growth curve created using the data of 2016 and 2017 after normalization. In step ST17, a growth curve is determined using data for a plurality of years after normalization. Thereby, a growth curve (universal curve) as shown in FIG. 9C is obtained.

(Verification of the universality of the growth curve)
The growth curve used for cultivation support is preferably a universal growth curve (universal curve) whose curve shape does not change depending on the year. The present inventors created growth curves for 2016 and 2017 using four types of explanatory variables using the size of the fruit subjected to normalization as an objective variable, and verified the universality of each. FIG. 10 is an explanatory diagram of a growth curve determined using four types of explanatory variables. FIG. 10 (a) is a growth curve with the number of days elapsed from April 1 as an explanatory variable, FIG. 10 (b) is a growth curve with an effective integrated temperature as an explanatory variable, and FIG. FIG. 10 (d) is a growth curve using the integrated value (integrated dimension) of the fractal dimension of the leaf region as an explanatory variable. Growth curves for 2016 and 2017 were created for any explanatory variable. When the integrated value of the leaf area is used as an explanatory variable, two growth curves with different fiscal years almost overlap, whereas when the other three types of indicators are used as explanatory variables, two different fiscal years are used. It can be seen that the growth curves of

  The two-year data used to verify the explanatory variables differed in weather conditions, with 2016 being a bad crop and 2017 being a good crop. When the integrated value of the leaf area was used as an explanatory variable, the degree of coincidence of the two growth curves used other explanatory variables, even though two growth curves were obtained using data with different weather conditions. Higher than the case. That is, if the integrated value of the leaf area is used as an explanatory variable, a universal growth curve that can be used regardless of the year can be obtained.

  The leaf area of this embodiment is obtained using a color image taken outdoors using sunlight as a light source. When a leaf region is obtained from a color image by the method described in Japanese Patent Application No. 2017-58688, the leaf region becomes small when the sunlight is shaded, and the leaf region becomes small when it is raining or cloudy. Means smaller. The present inventors inferred that this leaf area is related to nutrients generated by photosynthesis, and obtained the idea that the integrated value of the leaf area is used as an explanatory variable of a growth curve representing the growth of fruits. . As shown in FIG. 10, when the integrated value of the leaf area is used as the explanatory variable, the degree of coincidence of the growth curves of the multiple years is higher than when the other indicators are used as the explanatory variables, and it is related to the year. It was verified that a universal growth curve (universal curve) that can be used without failure was obtained. Therefore, the growth curve created using the integrated value of the leaf area as an explanatory variable was used for the evaluation of the growth state.

(Cultivation support service using growth curves)
FIG. 11 is an explanatory diagram of a cultivation support service using a growth curve. FIG. 11A shows a service method for providing advice on fruit-picking operations using a logistic curve formula, which is a growth curve, as a predictive formula for the fruit size (harvest diameter) at the time of harvest. The growth curve shown in FIG. 11 (a) is a logistic curve equation in which the size of fruit (apple) is normalized to a range of 0 to 1 and the coefficient a≈1, and here, 2016 in FIG. 9 (a). The formula is used.

In order to receive evaluation / advice using the growth curve, the data measurement unit 10 is installed in the farm of the farmer receiving the service, and color images of the fruit trees to be evaluated are taken according to a preset shooting schedule, and automatically every day. The integrated value of the leaf area is calculated in advance. From the prediction formula (growth curve) and the integrated value of the leaf area obtained by measurement at the farmer receiving the service, the growth state of the fruit is evaluated and advice is given. Evaluation and advice may be automatically performed according to a predetermined schedule, or may be performed at any time according to a request from a farmer who receives the service.

  For example, as shown in FIG. 11 (a), when the integrated value (integrated area) of the current leaf area is 4000, if the measured diameter of one of the growing fruits is 8.5 cm, the growth curve is used. From the prediction formula, it can be seen that the harvested diameter of the fruit is 1.000 / 0.654 times the current value, so it is about 13 cm. In this way, the size at the time of harvest of individual fruits can be predicted. Accordingly, the predicted harvest size can be provided as advice information.

  The integrated value of the current leaf area and the size at the time of harvest predicted from the current size of each fruit is the maximum size of this fruit when cultivation is continued as it is, and is the maximum size that can be harvested. If the harvested diameter is aimed to be 13 cm or more, the fruit whose current measured diameter is less than 8.5 cm is not expected to become 13 cm or more anymore, so all must be plucked. In this way, it is possible to quantitatively indicate the necessity of fruit removal and the size of fruit removal, and this information can be provided as advice information.

  FIG. 11B is a graph showing the relationship between the integrated value of the leaf area (integrated area) and the number of days elapsed. From the relationship shown by the prediction formula (growth curve), it is known that when the integrated value (integrated area) of the leaf area becomes 9700, the size (diameter 13 cm) can be harvested. From the transition of the integrated value (integrated area) of the leaf area up to now, the day when it reaches 9700 can be roughly predicted, so the harvest date can also be predicted. Therefore, the expected harvest date information can be provided as advice information. In apple cultivation, when the minimum temperature falls below a certain value, the texture of the apple decreases, so it is necessary to decide whether to prioritize increasing the size even if the texture is lowered or prioritizing the texture at the expense of size. Become. In order to make such a determination, the prediction of the harvest date is useful.

(Service for determining the necessity of additional fruit picking)
Next, a service will be described in which it is judged whether the farm to be serviced has grown in the same manner as a mature farmer, and work advice is provided if necessary. Fruit growth similar to that of an elderly farmer requires fruit removal work that keeps the balance of the number of leaves and fruits optimal. Ordinary farmers are generally unable to perform drastic fruit picking as they do in general. Therefore, in many cases, the proportion of fruits increases slightly. In that case, the actual fruit growth is a little slower than the slope of the growth curve. Therefore, the necessity of additional fruit picking is determined by comparing the slope of the approximate straight line that approximates the actual fruit growth and the slope of the tangent line of the growth curve (the ideal growth rate of the fruit).

  The approximate straight line approximating the growth of the fruit measured the transition of the radius of the fruit for 15 days for 7 days before and after the subject day as a single fruit, and measured the data for 15 days on the horizontal axis. Is determined by linear approximation with the integrated value of the leaf area. Then, the inclination of the determined approximate straight line is compared with the inclination of the tangent line of the growth curve on the target day. At this time, since the actual measurement data for 15 days is an actual size, the inclinations of the two cannot be compared as they are. Therefore, it is necessary to normalize the actual size. In this method, the ratio between the average value of the 15-day actual measurement data and the value of the growth curve (universal curve) on the target day is calculated, and the ratio is multiplied by the 15-day actual measurement data.

  If the slope of the approximate line is slower than the growth curve, additional fruit picking should be performed. On the other hand, if the slope is steeper than the growth curve, it is not necessary to take additional fruit after knowing it. In addition, even if it is a serious farmhouse, the judgment of the magnitude of this inclination has about +/- 8% dispersion | variation from the inclination of a growth curve (universal curve), Therefore If it is in the range of +/- 8%, additional work will be performed Absent. The above contents are provided as advice information.

When the above verification is performed, the transition of the radius of the apple for 15 days can be manually measured, but can also be performed automatically. The present inventors made AI (artificial intelligence) learn the process of automatically detecting fruits in a color image, and enabled automatic measurement. The general object detection algorithm used is called YOLO (You Only Look Once). When learning was performed by deep learning using data obtained by manually extracting fruits on about 3000 color images used for determining a growth curve as teacher data, AI was able to extract fruits satisfactorily.

  FIG. 12 is an explanatory diagram of automatic fruit detection by AI and automatic calculation of approximate straight lines. When the target date for obtaining the approximate line is the Nth day, the slope of the approximate line is obtained from the images for N ± p days. First, the fruit is automatically extracted from the image on the Nth day. As shown in FIG. 12A, each individual fruit (apple) extracted by AI is surrounded and displayed by a rectangle. Next, if the barycentric coordinates (x, y) and radius (r) of the extracted fruit are calculated from the output of the AI, a result as shown in FIG. 12B is obtained. Thus, the fruit position (center of gravity coordinates) and radius of the image on the target day (Nth day) are determined.

  Then, as shown in FIG.12 (c), about all the fruits in the image of a target day (Nth day), the same fruit is taken from the image measured on the day different from a target day (Nth day). Track and keep track of changes in radius. In this embodiment, the center of gravity ± radius range (x ± r) of images measured at the same time within ± p days with reference to the center of gravity (x, y) of the fruit on the target day (day N) for measuring the inclination. , Y ± r rectangular area), and from the found fruits, the fruit closest to the center of gravity was identified as the same fruit. In this embodiment, p = 7, and the same fruit is specified by searching for the range of the center of gravity ± the radius of the images for a total of 15 days of the subject date ± 7 days. When 8 or more fruits that can be identified as the same fruit as the Nth day were extracted from the 15-day image, linear approximation was performed using the least square method, and the inclination was calculated. However, before calculating the slope, the radius of the extracted fruit was normalized so that the square error between the radius of the extracted fruit and the growth curve was minimized. And the average value of all the calculated inclinations was calculated | required.

  FIG. 13 is an explanatory diagram of determination of necessity of fruit removal using an approximate straight line. FIGS. 13A to 13E show tangent lines and approximate straight lines of growth curves on different target days, respectively. Thus, for each target day, it is possible to visually determine whether or not fruit growth similar to that of a serious farmer has been achieved by simultaneously showing the slope of the tangent line of the growth curve and the slope of the approximate line. In FIG. 13, the ◯ marks indicate the radius of any one fruit whose inclination is calculated. Moreover, since the example of this figure uses the data of a serious farmer, the fluctuation is about ± 8% from the slope of the growth curve on any target day. In this form, if the calculated slope is -8% or more gentler than the growth curve, advice is given to perform additional fruiting, and if it is + 8% or more steep, it is indicated and added Advise that no work is required. Further, if the growth curve is within ± 8% of the inclination of the growth curve, it is in a good growth state, and advice is given that an additional fruit removal operation is unnecessary. In addition, since it is not possible to determine that the fruit extraction necessity determination using the approximate straight line does not elapse for 7 days after the target date, the inclination is automatically calculated from the data for the past 15 days at least every 7 days, It is desirable to give advice.

(Cultivation support system and cultivation support device)
Fig.14 (a) is a block diagram of the cultivation assistance system 1 to which this invention is applied. The cultivation support system 1 includes a data measuring unit 10 installed in a farm of a farmer who receives a service, a cloud server 20 (data storage unit) that can communicate with the data measuring unit 10 via a network such as the Internet, and a cloud server 20. And a cultivation support device 30 (data processing unit) that is a server capable of communicating via the Internet. The terminal 40 of the user who receives the service can communicate with the cultivation support device 30 and the cloud server 20 via the Internet. When receiving the cultivation support service, the cultivation support system 1 according to the present embodiment is premised on that the data measurement unit 10 is installed in each farm and the fruit tree image is automatically measured by automatic measurement. The data measuring unit 10 may include sensors for measuring environmental data such as air temperature, soil temperature, humidity, soil moisture, illuminance, and wind speed, and these sensors are controlled by the measurement computer 12 to obtain image data. At the same time, environmental data can be automatically measured. Data obtained by automatic measurement is accumulated in the cloud server 20. That is, the cloud server 20 functions as a data storage unit in the cultivation support system 1 to which the present invention is applied.

  FIG. 14B is a block diagram of the cultivation support device 30. The cultivation support device 30 functions as a data processing unit in the cultivation support system 1 to which the present invention is applied. The cultivation support device 30 is, for example, a general-purpose personal computer, and includes a display unit 31 such as a liquid crystal display, a computer main body 32, and an operation unit 33 such as a keyboard, a mouse, and a touch panel. The computer main body 32 includes a control unit 34 including a processing device such as a CPU, a storage unit 35 that stores various control programs including a cultivation support program, a communication unit 36 for connecting to the cloud server 20 via the Internet, and the like. Is provided. Note that the cultivation support device 30 and the cloud server 20 for storing image data can be integrated without being provided separately.

  As for the function of the data processing unit for executing the cultivation support method for performing the cultivation support service such as the evaluation using the growth curve and the provision of advice, the control unit 34 of the computer main body 32 stores the cultivation support program from the storage unit 35. This is realized by reading and executing. The cultivation support program causes the cultivation support apparatus 30 to perform a growth state evaluation process based on the growth curve and a process of determining advice information provided in the cultivation support service based on the growth state evaluation result. Further, the cultivation support program causes the cultivation support apparatus 30 to perform fruit size measurement processing using AI. The cultivation support device 30 may be a computer device other than a personal computer. That is, what is necessary is just to be provided with processing apparatuses, such as a non-volatile memory | storage means which memorize | stores a cultivation assistance program, and CPU which reads and runs a cultivation assistance program.

(Example of cultivation support method)
FIG. 15 is an explanatory diagram of an embodiment when the cultivation support method to which the present invention is applied is implemented in the cultivation support system 1 of FIG. A data measuring unit 10 is installed in the farm of the user of the cultivation support system 1 and automatic photographing is performed according to a schedule. The captured color image is sent from the data measuring unit 10 to the cloud server 20 and stored. The cultivation support device 30 that is a data processing unit calculates a leaf area from the color image acquired from the cloud server 20 and obtains an integrated value obtained by integrating the representative value of the leaf area every day. The obtained integrated value is stored in association with the image data. For an image containing fruit, the size of the fruit is automatically measured every day. The cultivation support device 30 performs the cultivation support service based on the evaluation result by evaluating the growth state according to a predetermined schedule or based on a request from the terminal 40.

  In addition to the cultivation support service, the cultivation support device 30 performs a monitoring service that displays accumulated color images and the like in time series based on a monitoring request from the terminal 40. For example, when a farmer who has installed a data measurement unit 10 in a plantation and becomes a user of the cultivation support system 1 logs in to the site of the cultivation support system 1, the color image of the plantation is displayed on the terminal 40 in time series. It is a service that can be displayed and monitor the growth state of plants in the farm. The cultivation support device 30 causes the terminal 40 to perform such display. The login information of the user of the cultivation support system 1 is associated with the color image photographed at the user's farm and the data (including the data for the cultivation support service) generated from the color image.

The service contents of the cultivation support service include size prediction at harvesting, advice on fruiting work, harvest date prediction, and advice on additional fruiting work, and other useful services can be added as appropriate. When the cultivation support apparatus 30 determines the advice, the cultivation support apparatus 30 notifies the user. The notification destination can be registered in advance. For example, the user's mail address can be notified. Moreover, when a user logs in to the site of the cultivation support system 1, advice is displayed on the user screen. The user can also record work information on the farm.

  Moreover, the cultivation assistance system 1 can provide services other than advice using monitoring and a growth curve. For example, the cultivation support apparatus 30 can manage user data (color images, advice information, work records) as shared information that can be browsed between a plurality of users. When browsing user data, a plurality of user data can be compared and displayed. As a result, the user can freely check the growth status, harvestable date, work record, etc. of the fruits of the farmers who belong to the same region, so that the cultivation support system 1 automatically presents the comparison with the status of his / her farm. You can consider whether to follow the advice you gave. When performing comparative display, statistical information such as averages and histograms can be displayed for those that can be quantified, and distribution information such as frequency when divided into several categories can be displayed for those that cannot be quantified. preferable. By doing so, it is possible to grasp the state of many crops and the collective knowledge of many users, and to make decisions with reference to them.

  The cultivation support system 1 can also provide functions such as a comment service, a work schedule sharing service, and a local community activation service as other services. The comment service is a function that accepts registration of comments on color images and advice information of other users and makes them available for viewing by the commented user or all users. The comment function enables advice and feedback between users, so that cultivation technology can be improved throughout the region. The work schedule sharing service is a function in which a user registers a work schedule of a farm and shares it with local farmers. If it is possible to know in advance the harvesting and shipping schedules of other farmers, it is possible to adjust the schedule, make effective use of manpower and materials, and adjust shipping. Local community activation services include event information announcements, participation receptions, and bulletin boards.

  In this way, the cultivation support system 1 can be extended to be operated as an agricultural community platform that provides services not only for individual farmers but also for the entire region. In Japan's agriculture, the persistence of individual farmers has become severe due to a decrease in the number of farmers and an aging population, and farming is often carried out while exchanging on a regional basis, mainly in regional JAs. Information exchange can be supported by providing an agricultural community platform.

  FIG. 16 is an example of a terminal 40 that uses the cultivation support system 1. A communication robot 41 that can communicate with the terminal 40 is introduced to enable the cultivation support system 1 to be used by conversation. Even if it is a farmer with low IT literacy, or even a person who is tired of farming work and loses the power to operate a personal computer, the work record by conversation and the data to the terminal linked to the conversation Presentation is possible.

(Main effects of this embodiment)
As described above, the cultivation support method, the cultivation assistance program, the cultivation assistance apparatus 30, and the cultivation assistance system 1 according to the present embodiment use a growth curve that approximates the growth state of fruits. Therefore, it is possible to appropriately grasp the current growing state of the fruit that directly affects the profits of the farmer. Moreover, the useful cultivation support service according to the present growth state of a fruit can be performed rather than providing the knowledge of a general cultivation method.

In this embodiment, the growth curve is determined with the fruit size as the objective variable and the integrated value of the leaf area as the explanatory variable. The leaf area is an area of a portion where photosynthesis is performed because the threshold value when binarizing the GBVI image (grayscale display) is calculated as a fixed value. Therefore, by using the integrated value of the leaf area, a growth curve that approximates the growth state of the fruit that grows by photosynthesis can be obtained. In addition, a universal growth curve that does not depend on environmental conditions such as weather can be obtained. Furthermore, after normalizing the fruit size data for each year, the growth curve is determined from the data for multiple years after normalization, so a universal growth curve (universal curve) that can be used regardless of the year is obtained. be able to. Therefore, even if there is a difference due to environmental conditions such as weather every year, useful cultivation support can be performed. For example, fruit growth prediction, harvest time prediction, and harvest size prediction can be performed. Moreover, it is possible to evaluate the fruit picking work. Furthermore, if the growth curve changes despite the proper operation, there may be a cause of disease or pest. Since the occurrence of diseases, pests and the like can be confirmed visually, if such a situation occurs, provision of useful information using this growth curve can be avoided.

  In this embodiment, since the growth curve is determined based on the data of the size of the fruit and the integrated value of the leaf area in the fruit tree cultivation performed by the successful farmer, the fruit being grown is brought close to the same growing state as the successful farmer. Cultivation support can be performed. Therefore, the profit of the farmer can be increased.

  In this embodiment, when performing the cultivation support service, the data measuring unit 10 is used to construct a system that automatically captures an image of the user's farm and automatically stores the image in the cloud server 20. The cultivation support device 30 (data processing unit) automatically calculates the integrated value of the leaf area, automatically measures the size of the fruit from the image, automatically evaluates the growth state, and automatically determines the advice. And notify. Therefore, the labor for carrying out the cultivation support service can be reduced, and the burden on the user can be reduced. Also, advice information can be provided to the user at an appropriate timing.

  In this embodiment, the cultivation support system 1 has various additional functions. For example, the growth state can be monitored. Further, growth state data, advice information, work record data, and the like can be shared among users, and data can be compared and displayed. Accordingly, each user can grasp the state of many crops and the collective knowledge of many users, and can make a decision with reference to them. Furthermore, it is possible to provide a comment function for growth state data and advice, and a function for sharing the work schedule of each user. In addition, a community community activation service function can be provided. Therefore, it can be operated as an agricultural community platform that provides services not only for individual farmers but also for the entire region.

(Other forms: Cultivation support method for crops in general)
The above form is a method for evaluating the growth state of a fruit by using a growth curve that approximates the growth state of a fruit by cultivation of a serious farmer. The definition of (area) is not the area of the actual leaf, but the area of the portion that is exposed to light and performing photosynthesis. Therefore, the present invention can be applied to agricultural crops other than fruit trees. For example, with respect to rice, the growth state of rice can be evaluated by determining a growth curve with the integrated value of the leaf area as the horizontal axis and the size of the ear of rice that is the harvest target portion as the vertical axis. Further, regarding leaf vegetables, the growth state of leaf vegetables can be evaluated by determining a growth curve with the integrated value of the leaf area as the horizontal axis and the size of the portion to be harvested as the vertical axis. Thus, the present invention can be applied to all crops.

FIG. 1B is a flowchart of a cultivation support method for all agricultural products, which is the broadest concept of the present invention. As shown in FIG.1 (b), in the cultivation assistance method of this invention, the growth curve which approximated the growth state of the harvesting target part in agricultural products is determined previously (step ST1a). Next, the growth state of the portion to be harvested in the crop being grown is evaluated using the growth curve (step ST2a). And based on the evaluation result of a growth state, the cultivation assistance service regarding the crop under cultivation is performed (step ST3a).

  In this specification, the cultivation support service is to provide advice information to the user. The advice information may be anything useful for the user (farmer). For example, the advice information is information useful for cultivating according to the target, and includes work proposals such as what work should be done at what time and how much. Further, the predicted result such as the yield of the harvest, the size at the time of harvest, and the harvest time may be used. Or the evaluation result of a growth state can also be provided as advice information as it is.

  The growth curve determined in step ST1a is a curve that approximates the growth state of the portion to be harvested in the crop. The cultivation support method of the present invention aims at realizing a growing state achieved by tacit farmers by tacit knowledge such as cans and experiences for general farmers. Therefore, the growth curve of step ST1a is determined using data on the growth state of the crop cultivated by a serious farmer. Or the data of the growth state at the time of growing so that a specific objective may be implement | achieved, such as harvesting the fruit of a specific size, can also be used.

DESCRIPTION OF SYMBOLS 1 ... Cultivation support system, 10 ... Data measurement part, 11 ... Camera, 12 ... Computer for measurement, 13 ... Communication part, 14 ... Power supply part, 20 ... Cloud server, 30 ... Cultivation support apparatus, 31 ... Display part, 32 ... Computer main body 33 ... operation part 34 ... control part 35 ... storage part 36 ... communication part 40 ... terminal 41 ... communication robot

Claims (25)

Using the growth curve that approximates the growth state of the harvested portion of the crop, evaluate the growth status of the harvested portion of the crop being grown,
The cultivation support method characterized by performing the cultivation assistance service regarding the said agricultural crop under cultivation based on an evaluation result.   2. The growth curve according to claim 1, wherein the growth curve is determined using the size of the portion to be harvested as an objective variable, and an integrated value of a leaf area that is an area of a leaf area in which the crop is captured as an explanatory variable. The cultivation support method as described. The crop is a fruit tree, and the harvest target portion is a fruit,
The cultivation support method according to claim 2, wherein the growth curve is determined using the size of the fruit as an objective variable and the integrated value of the leaf area as an explanatory variable.   The cultivation support method according to claim 3, wherein the growth curve is determined based on an integrated value of the size of the fruit and the leaf area in fruit tree cultivation performed by a serious farmer. Collect the data of the fruit size and the integrated value of the leaf area over multiple years,
The cultivation according to claim 3 or 4, wherein the growth curve is a universal curve obtained from the data of the plurality of years after normalization after normalizing the data of the fruit size for each year. Support method.   6. The cultivation support according to claim 3, wherein the integrated value of the leaf area is a value obtained by integrating the maximum value of the leaf area obtained during one day every day. Method.   The cultivation support method according to any one of claims 3 to 6, wherein the growth curve is a logistic curve.   From the image of the fruit tree being cultivated, find the integrated value of the leaf area up to the point of time when the growth state is evaluated, and based on the value on the growth curve corresponding to the integrated value, in the fruit tree being cultivated The cultivation support method according to any one of claims 3 to 7, wherein a growth state of the fruit is evaluated.   The cultivation support method according to any one of claims 3 to 8, wherein the cultivation support service is a size prediction of the fruit at the time of harvest.   The cultivation support method according to any one of claims 3 to 9, wherein the cultivation support service is an advice on a fruit picking operation.   The cultivation support method according to any one of claims 3 to 10, wherein the cultivation support service is prediction of a harvest date of the fruit. Measure the size of the fruit in the fruit tree during cultivation for a predetermined period, and obtain an approximate straight line approximating the change in the size of the fruit,
Compare the slope of the approximate line with the slope of the tangent line of the growth curve,
The cultivation support method according to any one of claims 3 to 11, wherein advice on additional fruit removal work is performed based on the comparison result. The cultivation support method according to claim 12, wherein the size of the fruit is measured by extracting the fruit from a color image obtained by photographing the fruit tree.   The cultivation support method according to claim 13, wherein the measurement of the fruit size is performed using an AI that has learned a process of extracting the fruit using a color image obtained by photographing the fruit tree. For each single pixel or unit region of the color image, using a G image that is an image in the green wavelength region of the color image obtained by photographing the crop and a B image that is an image in the blue wavelength region of the color image, GBVI is calculated by the formula (A) of
GBVI = (GB) / (G + B) (A)
For each single pixel or unit region, generate a GBVI image obtained by binarizing the pixel value generated from the GBVI based on a threshold value that is a fixed value;
The cultivation support method according to any one of claims 2 to 14, wherein the leaf area is obtained from the GBVI image.   The computer provides to the cultivation support service based on the evaluation process of the growth state of the crop under cultivation based on the growth curve in the cultivation support method according to any one of claims 1 to 15, and based on the evaluation result. The cultivation support program characterized by performing the process which determines advice information to be performed.   The cultivation support program according to claim 16, which causes a computer to perform measurement processing of the fruit size in the cultivation support method according to claim 13 or 14. A storage unit for storing the cultivation support program according to claim 16 or 17,
And a control unit that reads and executes the cultivation support program from the storage unit. A camera that captures color images with visible light;
A data storage unit for storing the color image photographed by the camera;
A measurement unit that controls the camera to shoot the color image in a predetermined shooting schedule, and stores the color image in the data storage unit;
The integrated value of the leaf area in the cultivation support method according to any one of claims 2 to 15 is calculated using the color image read from the data storage unit, and the integrated value of the leaf area, A data processing unit for determining advice information provided in the cultivation support service using the growth curve in the cultivation support method according to any one of Items 2 to 15, and a cultivation support characterized by system. The data processing unit is provided in a server connected to a network,
The cultivation support system according to claim 19, wherein the server notifies the cultivation support service to a notification destination registered in advance via the network.   Based on a request from a terminal connected to the server via the network, the server uses the color image stored in the data storage unit to display a display for monitoring a growing state of a crop. The cultivation support system according to claim 20, wherein the cultivation support system is performed.   The server manages all or a part of the color image and the advice information associated with each of a plurality of users as shared information that can be browsed by all or a part of the plurality of users. The cultivation support system according to claim 20 or 21 characterized by things.   The cultivation support system according to claim 22, wherein the server compares and displays all or part of the color image and the advice information associated with each of a plurality of users.   The cultivation support system according to claim 22 or 23, wherein the server receives a registration of a comment on the color image and the advice information and manages the comment.   The cultivation support system according to any one of claims 22 to 24, wherein the server receives registration of a work schedule from each of the plurality of users and manages the work schedule as shared information.