WO2024018559A1

WO2024018559A1 - Shipping quantity prediction method and shipping quantity prediction system

Info

Publication number: WO2024018559A1
Application number: PCT/JP2022/028196
Authority: WO
Inventors: 由久宇佐美; 正裕北島
Original assignee: 株式会社ファームシップ
Priority date: 2022-07-20
Filing date: 2022-07-20
Publication date: 2024-01-25

Abstract

Provided are a shipping quantity prediction method and a shipping quantity prediction system for predicting the shipping quantity of crops to be shipped from fields in an entire region to a market. This shipping quantity prediction method comprises: a step for identifying the harvest timing of crops in a target region, in which the crops are planted, on the basis of an image obtained by photographing from the sky the target region; a step for calculating the growth rate of the crops on the basis of the harvest timing; a step for predicting a yield of the crops in the target region at the harvest timing on the basis of the growth rate having been calculated currently and growth rates in the past; and a step for estimating the shipping quantity of crops to be shipped from a plurality of the target regions to the market on the basis of yields predicted from the respective target regions.

Description

Shipping volume prediction method and shipping volume prediction system

The present invention relates to a shipping amount prediction method and a shipping amount prediction system using images taken from above of target areas where crops are planted, and in particular, to estimate the amount of crops shipped to the market from multiple target areas. The present invention relates to a shipping amount prediction method and a shipping amount prediction system.

Traditionally, crop yields have been predicted using various methods. For example, a surveyor goes to the site where crops are planted, observes the growing conditions of the crops, and predicts the yield based on experience. It takes time for researchers to predict crop yields. In particular, when the cultivated area is large, it takes more time and costs to predict the yield. In addition, predictions of yields based on the experience of surveyors vary from surveyor to surveyor, resulting in variations in yield predictions, and as a result, the reliability of yield predictions is poor. In addition to predicting crop yields by researchers, for example, a harvest prediction device is proposed in Patent Document 1.

The harvest prediction device of Patent Document 1 is a harvest prediction device that predicts the yield of a crop from image data of the crop. It stores growth stage spectrum information, which is information about the crop, growth period information, which is information about the period from the start of crop growth to harvest, and area yield information, which is information about the relationship between the planted area of the crop and the yield. A crop in which the type of crop is identified by comparing information regarding the spectrum obtained from the image data with the type spectrum information stored in the storage unit, and crop type information that is information regarding the identified type is obtained. The type identification unit identifies the growing stage of the crop by comparing the spectrum information obtained from the image data with the growing stage spectrum information stored in the storage unit, and the growing stage is information about the identified growing stage. A growing stage identification unit that obtains information, a planted area calculation unit that calculates a planted area based on image data and obtains planted area information that is information about the calculated planted area, and crop type information, growing stage information, and planted area information. and a harvest prediction section that predicts the harvest time and yield of crops from the growing period information and area yield information stored in the storage section.

In addition to Patent Document 1, for example, Patent Document 2 describes at least one planting date for a crop and a planting amount, which is the amount of the crop to be planted on the planting date, as a method for harmonizing the demand amount and the harvest amount of the crop. A cropping schedule calculation device that calculates the amount has been proposed. In Patent Document 2, the harvest amount of crops is brought closer to the demand amount.

JP2003-006612A Japanese Patent Application Publication No. 2021-002110

If there are multiple fields in a specific region, and if the harvest time and yield of crops are predicted for each field using Patent Document 1, for example, the prediction of the harvest time and yield of crops in each field I can. However, it is not possible to predict the amount of crops shipped from fields throughout the specific region mentioned above.
Further, if the above-mentioned Patent Document 2 is used to bring the harvest amount of crops closer to the demand amount, it is possible to bring the harvest amount of crops in each field closer to the demand amount. However, it is not possible to predict the amount of crops shipped from fields throughout the specific region mentioned above.
If it is not possible to predict the amount of crops shipped from fields in a particular region as mentioned above, many of the same type of crops may be shipped to the market, resulting in an oversupply of crops, or conversely, there may be a shortage of crops, causing the market to decline. It is not possible to stably ship a certain amount of crops.
For this reason, it is desired to predict the amount of crops shipped from fields throughout the region to the market.

An object of the present invention is to provide a shipment amount prediction method and a shipment amount prediction system that predict the shipment amount of crops shipped from fields in the entire region to the market.

In order to achieve the above-mentioned object, the invention [1] includes a step of identifying the harvest time of crops in a target area based on an image taken from above of the target area where the crops are planted, and A process of calculating the growth rate of crops, a process of predicting the yield of crops in a target area at the time of harvest based on the currently calculated growth rate and past growth rates, and a process of predicting the yield of crops in a target area at the time of harvest, and each of a plurality of target areas. The present invention provides a method for predicting the amount of shipments, which includes the step of estimating the amount of shipments to markets where crops are shipped from a plurality of target areas, based on the predicted yields.

In the invention [2], in the process of identifying the harvest time of crops in a target area, a current image taken from above of the target area where crops are planted and a past image taken from above of the target area where crops were planted are used. The shipping amount prediction method according to invention [1] calculates the color distance to the image and specifies the harvest time of crops in the target area based on the color distance.
Invention [3] obtains in advance the correlation between the crop yield in the target area at the past harvest time and the growth rate, and in the step of predicting the crop yield in the target area at the harvest time, the correlation is obtained. The shipping amount prediction method according to [2] or [2], which uses the relationship to predict the harvest amount in the target area at the time of harvest.
Invention [4] is the shipping amount prediction method according to any one of inventions [1] to [3], wherein the crop is a leafy vegetable.

Invention [5] includes a specific unit that identifies the harvest time of crops in a target area based on an image taken from above of the target area where the crops are planted, and a calculation unit that calculates the growth speed of the crops based on the harvest time. a prediction unit that predicts the crop yield in the target area at the time of harvest based on the currently calculated growth rate and past growth rates; , and an estimator that estimates the amount shipped to a market where crops are shipped from a plurality of target regions.

In the invention [6], the identification unit calculates the color distance between a current image taken from above of a target area where crops are planted and a past image taken from above of a target area where crops are planted, The shipping amount prediction system according to invention [5] specifies the harvest time of crops in a target area based on color distance.
Invention [7] obtains in advance the correlation between the yield of crops in the target area at the time of past harvest and the growth rate, and the prediction unit uses the correlation to predict the yield in the target area at the time of harvest. The shipping amount prediction system according to invention [5] or [6] predicts the yield of crops.
Invention [8] is the shipping amount prediction system according to any one of inventions [5] to [7], wherein the crop is a leafy vegetable.

According to the present invention, it is possible to provide a shipping amount prediction method for predicting the shipping amount of crops shipped from fields in the entire region to the market.
Further, it is possible to provide a shipping amount prediction system that predicts the shipping amount of crops shipped from fields in the entire region to the market.

FIG. 2 is a schematic diagram showing an example of a region targeted by a shipping amount prediction method according to an embodiment of the present invention. FIG. 1 is a schematic diagram showing an example of a shipping amount prediction system according to an embodiment of the present invention. 3 is a flowchart illustrating an example of a shipping amount prediction method according to an embodiment of the present invention. (a) to (d) are schematic diagrams for explaining an example of a method for specifying a planting time and a harvest time in a shipping amount prediction method according to an embodiment of the present invention. (a) and (b) are schematic diagrams for explaining an example of a method for specifying the harvest time of crops in the shipment amount prediction method according to the embodiment of the present invention. It is a flowchart which shows an example of the process of calculating the growth rate of the shipping amount prediction method of embodiment of this invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The shipping amount prediction method and shipping amount prediction system of the present invention will be described in detail below based on preferred embodiments shown in the accompanying drawings.
Note that the figures described below are illustrative for explaining the present invention, and the present invention is not limited to the figures shown below.
In addition, in the following, "~" indicating a numerical range includes the numerical values written on both sides. For example, when ε is a numerical value ε _α to a numerical value ε _β , the range of ε is a range that includes the numerical value ε _α and the numerical value ε _β as upper and lower limits, and expressed in mathematical symbols as ε _α ≦ε≦ε _β. be.

FIG. 1 is a schematic diagram showing an example of a region targeted by the shipping amount prediction method according to the embodiment of the present invention. FIG. 1 schematically shows a production area 10 in order to explain a shipping amount prediction method and a shipping amount prediction system. The production area 10 is an area where crops are cultivated, and the target area is a cultivation area within the production area 10. The unit (section unit) that defines the cultivation area may be a unit at the level of each prefecture, a unit at the municipal level, or a unit smaller than the municipality.

The production area 10 includes a plurality of cultivation areas, and the production area 10 shown in FIG. 1 includes, for example, three cultivation areas 10a, 10b, and 10c. These three cultivation areas 10a, 10b, and 10c correspond to target areas, respectively.
Furthermore, the cultivation area 10a includes, for example, four fields 11a. Furthermore, in the cultivation area 10b, there are, for example, 12 fields 11b. Furthermore, the cultivation area 10c has 32 fields 11c. A field is a field where crops are grown. Note that there are no particular restrictions on the scale, planted area, etc. of each field in each cultivation region, and the fields may be arranged regularly or irregularly in each cultivation region.
In the fields 11a, 11b, and 11c, crops are planted and harvested after growing to a predetermined size.

The crops are, for example, leafy vegetables. Leafy vegetables are also called leafy vegetables. Examples of leafy vegetables include lettuce, komatsuna, cabbage, spinach, broccoli, garland chrysanthemum, bok choy, Chinese cabbage, mizuna, chive, and green onion.

(Shipping volume prediction system)
Next, the shipping amount prediction system will be explained.
FIG. 2 is a schematic diagram showing an example of a shipping amount prediction system according to an embodiment of the present invention. The shipping amount prediction system 20 shown in FIG. 2 is an example of a system used in the shipping amount prediction method, and is configured using, for example, hardware such as a computer or software on the cloud. As long as the shipping amount prediction method can be executed using hardware and software such as a computer, the shipping amount prediction method is not limited to using the shipping amount prediction system 20 shown in FIG. 2. That is, the shipping amount prediction method of the present invention may be implemented using a configuration (system) other than the shipping amount prediction system 20 of FIG. 2. Further, a program for causing a computer or the like to execute each step of the shipping amount prediction method may be used.

The shipping amount prediction system 20 shown in FIG. 2 includes, for example, a processing section 22, an input section 24, and a display section 26. The calculation unit 32 includes an acquisition unit 30 , a specification unit 31 , a calculation unit 32 , a prediction unit 33 , an estimation unit 34 , a display control unit 35 , a memory 36 , and a control unit 37 . Although not shown, the shipping amount prediction system 20 also includes a ROM and the like.
The processing section 22 is controlled by a control section 37. Further, in the processing unit 22 , the acquisition unit 30 , the identification unit 31 , the calculation unit 32 , the prediction unit 33 , the estimation unit 34 , and the display control unit 35 are connected to the memory 36 . Furthermore, the data output from each of the acquisition section 30 , identification section 31 , calculation section 32 , prediction section 33 , estimation section 34 , and display control section 35 can be stored in the memory 36 .

The shipping amount prediction system 20 has the control unit 37 execute a program (computer software) stored in a ROM or the like, thereby controlling the acquisition unit 30, the identification unit 31, the calculation unit 32, the prediction unit 33, and the estimation unit 34. Form each part functionally. As described above, the shipping amount prediction system 20 may be configured by a computer in which each part functions by executing a program, or may be a dedicated device in which each part is configured with a dedicated circuit. It may be configured on a server to run on the cloud.

The shipment amount prediction system 20 was constructed for the purpose of predicting the shipment amount of crops to be shipped to the market from data related to fields in the entire region, especially image data of each field. Each part of the shipping amount prediction system 20 will be explained below.

The input unit 24 is various input devices such as a mouse and a keyboard for inputting various information according to instructions from an operator.
The display section 26 displays, for example, the predicted shipment amount of crops to be shipped from fields in the entire region to the market, and various known displays are used. The display unit 26 also includes a device such as a printer for displaying various information on an output medium.

In the acquisition unit 30, for example, an image taken from above of a target area where crops are planted is inputted from outside the shipping amount prediction system 20 in the form of image data by the input unit 24, and is held therein. . Image data of an image taken from above of the target area where the above crops are planted, which is held in the acquisition unit 30 , is read out to the identification unit 31 . Note that the above-described image data held in the acquisition unit 30 may be stored in the memory 36.

It is preferable that the image of the target area where crops are planted be taken from above, for example, once a day, and more preferably, once a day, at the same time. Further, it is preferable that the image includes shooting date information and location information.
Level correction, atmospheric correction, geometric correction, etc. may be performed on the image data of an image taken of the target area from above. Level correction is correction that removes variations in sensitivity or noise of the sensor that acquired the image data. In addition, atmospheric correction is correction that removes noise caused by water vapor or the like when image data is acquired. In addition, geometric correction is correction that removes geometric distortion and the like included in an image.

Note that the image data of the image taken from above of the target area where crops are planted is not particularly limited as long as it is computer-readable data and computer-analyzable data. Further, the acquisition unit 30 may convert image data of an image taken from above of a target area where crops are planted into a data format that can be processed by the identification unit 31.
As will be described later, in order to utilize information on a specific wavelength of the image data, it is preferable that the image data of an image of the target area taken from above includes information on the specific wavelength. Specific wavelengths will be explained later.
Here, the target areas where crops are planted are, for example, three cultivation areas 10a, 10b, and 10c of the production area 10 shown in FIG.

Images taken of the target area from above are not particularly limited, and can be taken and obtained using aircraft such as airplanes and helicopters, artificial satellites, or drones. In this case, it is preferable that the photographed image includes positional information based on the photographing position. This makes it easier to identify the field in the target area in the photographed image by comparing the position information of the photographed image with the position information of the target area.

The specifying unit 31 specifies the harvest time of crops in the target area based on an image (image data) taken from above of the target area where the crops are planted.
The harvest time of crops in the target area refers to a state in which there are no crops in the field, such as when the crops have been harvested from the field, and is the harvest date and time. In a field, the appearance of the soil from above changes depending on the presence or absence of crops and the growth conditions of the crops at the time of planting or harvesting. The images (image data) are different.
The specifying unit 31 uses the above-mentioned differences in images (image data) to specify the harvest time.
The identification unit 31 calculates the color distance between the current image taken from above of the target area where crops were planted and the previous image taken from above of the target area where crops were planted, and then calculates the color distance based on the color distance. Identify the harvest time of crops in the target area.
The past image is an image of the same area (field) as the current image, and is used to determine the growth status of crops in the current image or whether the crops in the current image have already been harvested.

As a procedure for identifying the harvest time using the current image and past images, for example, when the image data of the image is expressed in 8 bits and 256 gradations, if the color distance changes by 10 or more, the harvest time shall be. Therefore, in order to identify the harvest time, it is preferable to take an aerial image of the target area where crops are planted once a day.
For the color distance, at least one wavelength of the image data may be used, but it is preferable to use a plurality of wavelengths.
The specifying unit 31 causes the memory 36 to store the information on the time of harvest of the crops described above and the information on the color distance.

The calculation unit 32 calculates the growth rate of crops based on the harvest time.
The growth rate is, for example, the number of days from planting to harvesting. For this reason, it is necessary to specify the planting date (time of planting).
Planting is when crops are planted in fields where there are no crops. The planting time can be identified by color distance, similar to the above-mentioned harvesting time.
In fixed planting, when the image data of an image is expressed in 8 bits and 256 gradations, if the color distance changes by 10 or more, the photographing date and time of the image is taken as the fixed planting time. Therefore, in order to identify the planting time, it is preferable to take an aerial image of the target area where crops are planted once a day.
When specifying the planting time, an image of the field before planting to be specified (in which the soil is exposed) is taken in advance. The color distance between the image of the field before planting to be identified and the photographed image of the field to be identified is determined. When this color distance is 10 or more, the date and time when the image of the field was taken is defined as the planting time.
Note that the growth rate will be explained in detail later.

The prediction unit 33 predicts the yield of crops in the target area at the time of harvest based on the currently calculated growth rate and past growth rates.
For prediction purposes, the correlation between crop yields and growth speeds in the target area at past harvest times is obtained in advance. As information based on the correlation, for example, the yield amount when the number of days from planting to harvest date is 50, 70, and 100 days is obtained in advance for each type of crop or for each field. The number of days from the planting date to the harvesting date represents the growth rate of the crop (more specifically, the past growth rate).
The correlations obtained in advance for each type of crop or for each field are stored in the memory 36, for example. The prediction unit 33 reads the above-mentioned correlation from the memory 36 and predicts the yield of crops in the target area at the time of harvest. Specifically, based on the above-mentioned correlation, the calculation unit 32 predicts the harvest amount corresponding to the currently calculated growth rate.
The prediction unit 33 predicts the harvest amount at the time of harvest in each of the plurality of target areas. Specifically, the yields of three cultivation areas 10a, 10b, and 10c of the production area 10 shown in FIG. 1 are predicted. That is, the prediction unit 33 predicts the yields of the fields 11a, 11b, and 11c shown in FIG. 1, respectively.

The method of determining the correlation is not particularly limited, and various known methods can be used. As the correlation, for example, Pearson's correlation analysis or covariance is used.

The estimating unit 34 estimates the amount of crops shipped from the plurality of target regions to the market based on the predicted harvest amount for each of the plurality of target regions.
Specifically, using the yields of the three cultivation areas 10a, 10b, and 10c of the production area 10 shown in FIG. presume. Crops harvested in each of the 10 growing regions (more specifically, each field in each growing region) are shipped to the market.
As described above, the harvest amount is predicted for each field in each cultivation region based on the specified harvest date. Therefore, if the period from the harvest date to the time of shipment to the market is known, it is possible to know on what date and how much of the crops harvested from the field will be shipped to the market. The estimation unit 34 uses the harvest amount predicted by the prediction unit 33 in the situation where the harvest date is specified for each of the fields 11a, 11b, and 11c in the three cultivation areas 10a, 10b, and 10c, and calculates the specific date. It is possible to estimate the amount of crops shipped to the market from the 10 production areas in .
The amount of crops shipped to the market estimated by the estimation unit 34 is stored in the memory 36, for example.

The display control unit 35 uses the image data acquired by the acquisition unit 30, the information on the harvest time specified by the identification unit 31, the growth rate calculated by the calculation unit 32, and the crops in the target area predicted by the prediction unit 33. The harvest amount and the amount shipped to the market estimated by the estimation section 34 are displayed on the display section 26.
In the display control section 35, when displaying on the display section 26, various information may be read out from the memory 36 and displayed. Further, the display control section 35 can also display various information inputted via the input section 24 on the display section 26. Further, the information to be displayed on the display section 26 may be transmitted using a transmitting section (not shown).

(Shipping amount prediction method)
FIG. 3 is a flowchart showing an example of a shipping amount prediction method according to an embodiment of the present invention.
In the shipment amount prediction method, first, the harvest time of crops in the target area is specified based on an image taken from above of the target area where the crops are planted (step S10).
For example, images are obtained by photographing the three cultivation areas 10a, 10b, and 10c of the production area 10 shown in FIG. 1 from above using an aircraft, an artificial satellite, or a drone. Note that the above-mentioned image acquisition frequency is, for example, once a day.
The image data of the obtained image is output to the acquisition unit 30 of the processing unit 22 of the shipping amount prediction system 20, for example.

An image taken from above includes, for example, RGB, NIR, and UV as wavelength components.
RGB are the three primary colors used in regular cameras. There are various choices for the wavelengths of RGB, but typically R has a wavelength of 600 to 700 nm, G has a wavelength of 500 to 600 nm, and B has a wavelength of 400 to 500 nm.
NIR is called near infrared rays and has a wavelength of 700 to 800 nm.
UV is called ultraviolet light and has a wavelength of 200 to 400 nm.

In addition to normal RGB images, images taken from above include images taken by a multispectral camera, images taken by an infrared camera or an ultraviolet camera, and images taken by reflection of radio waves, and the images are not particularly limited. It is preferable to use images with red (R) wavelengths or near-infrared wavelengths (NIR) because plants can be easily detected. For this reason, it is preferable for an image taken from the sky to have only red (R) or only red (R) and near-infrared wavelengths (NIR) as wavelength components.
For this reason, the camera (imaging device) that captures images taken from above is not particularly limited as long as it has sensitivity to the above-mentioned wavelengths, but it is preferably one that can capture a wide range at once. For this reason, it is preferable to use a camera mounted on the above-mentioned artificial satellite. Note that the number of cameras (imaging devices) is not particularly limited, and may be one or more.

As described above, the harvest time of crops in the target area is determined based on the image taken from above (step S10).
Here, FIGS. 4(a) to 4(d) are schematic diagrams for explaining an example of a method for specifying the planting time and harvesting time in the shipping amount prediction method according to the embodiment of the present invention.
4(a) shows a field 40 before planting crops, FIG. 4(b) shows a field 41 with crops 13 planted, and FIG. 4(c) shows a field 42 during the harvest period of crops 13. FIG. 4(d) shows a field 43 after crops are harvested.
As shown in FIGS. 4(a) to 4(d), in the fields 40 to 43, the degree of soil exposure, etc. is different before the planting of crops, in the state where the crops have been planted, during the harvest period of the crops, and after the harvest of the crops. different. Utilizing this, for example, the color distance can be calculated from an image taken from the sky, and based on the color distance, the state of the target area before planting, the state of no crops after harvesting, and the state of no crops after the crops have been planted. identify the conditions in which crops are harvested, as well as when crops are harvested.

For example, FIGS. 5A and 5B are schematic diagrams for explaining an example of a method for specifying the harvest time of crops in the shipment amount prediction method according to the embodiment of the present invention. 5(a) and (b) show the cultivation area 10b shown in FIG. 1. The cultivation area 10b has 12 fields 11b as described above.
For example, position information has been acquired for the 12 fields 11b in the cultivation area 10b. Further, the image (image data) of the field 11b, which will be described later, includes position information. Thereby, it is possible to identify which of the 12 fields 11b in the cultivation area 10b is in each photographed image.
In the cultivation area 10b, an image (image data) of the field 11b shown in FIG. 5(a) is obtained before planting. When the image (image data) of the field 11b shown in FIG. 5(a) is obtained continuously every day, there is no change in the image (image data) before planting.
On the other hand, for example, if crops 13 are planted in some fields 12a of the fields 11b shown in FIG. 5(a) as shown in FIG. 5(b), an image of the fields 12a where the crops 13 are planted (image data) changes. Changes in this image (image data) are detected using, for example, color distance.

For example, the color distance between a current image taken of the target area from above and a past image of the target area taken from above is used for the change in the image (image data). Here, the past image may be an image taken one day or more before the image to be compared (that is, the current image), but since planting the crop 13 does not require much time, past images may be used. As long as it is an image from one day ago. Past images are stored in the memory 36, for example.
For example, color data of a specific area of an image (image data) is used for the color distance. The specific area will be explained later, but the specific area is, for example, the area in the middle of the photographed field.
Further, as the past image, an image obtained by averaging images from a normal year may be used. Further, as the past image, normal color data including color data before planting in a normal year and color data immediately before harvest in a normal year, which will be described later, may be used.

The image data of the image of the field 12a in FIG. 5(a) is designated as D1, and the image data of the image of the field 12a in which the crops 13 are planted in FIG. 5(b) is designated as D2. Note that both the image data D1 and the image data D2 are represented by 8 bits and 256 gradations, for example, including the image data values of gradation 0 or gradation 255.
When specifying the planting date using the wavelength component R for the field 12a in FIGS. 5(a) and (b), the wavelength component R1 of R in image data D1 and the wavelength component R2 of R in image data D2 If the color distance is 256 gradations and the distance is 10 or more, it is determined that the images have been planted. In this way, the planting date can be specified based on the color distance calculated from the image data of each date and time. The specified planting date is stored in the memory 36. Note that when the color distance is Dc, Dc is expressed by the following formula.

In addition to the wavelength component R described above, the color distance may be a combination of the wavelength component R and NIR. In this case, the color distance Dc is expressed by the following formula.

The color distance may also be a combination of wavelength components R, G, and B. In this case, the color distance Dc is expressed by the following formula.

The procedure for specifying the planting date has been described above, but the procedure for specifying the harvest time is also the same as the procedure for specifying the above-mentioned planting date. The color distance between the current image taken from above of the target area where crops were planted and the past image taken from above of the target area where crops were planted is calculated, and the color distance of the crops in the target area is calculated based on the color distance. Identify the point of harvest.
Specifically, for example, the image data Ds of the field 41 shown in FIG. 4(b) and the image data Dh of the field 42 shown in FIG. 4(c) are compared to identify the harvest time. In this case, as described above, the color distance between the image data Ds and the image data Dh is determined to specify the harvest time. Regarding the color distance, as described above, it may be only R, a combination of R and NIR, or a combination of R, G, and B. If the color distance is 256 gradations and the distance is 10 or more, it is determined that crops have already been harvested in the target area in the current image. Then, based on the determination result, the harvest time is specified based on the date and time when the current image was taken. The identified harvest time is stored in the memory 36, for example, as a harvest date.

Next, the growth rate of the crop is calculated based on the harvest time (step S12). In step S12, for example, a method shown in FIG. 6 described below is used. FIG. 6 shows an example of step S12, which is executed by the calculation unit 32 of FIG. 2, for example.

<How to calculate crop growth rate>
FIG. 6 is a flowchart showing an example of the process of calculating the growth rate of the shipping amount prediction method according to the embodiment of the present invention.
In calculating the growth rate of crops in the target area (step S12), first, a specific field is photographed from above on a specific day to obtain an image, and image data is obtained (step S20). The specific date in step S20 is the photographing date of the image. As described above, it is preferable that the image includes shooting date information and location information.

Next, color data of a specific area of the image is acquired (step S22).
The specific area in step S22 is the area in the middle of the field in the photographed image. Among the image data of the middle area of the field, the color data of the wavelength component R, the wavelength component R and NIR, and the wavelength component RGB described above are used. In the case of wavelength components R and NIR, the average value of R and NIR is used as color data. In the case of wavelength components RGB, the average value of RGB is used as color data.
The center of the field is the geometric center determined from the outline of the field in the image (image data). The outline of the field can be identified by processing the image data of the photographed image and extracting the outline of the field.

Next, normal year color data is acquired from the database (step S24).
The normal color data is the average value of the color data of the middle area (specific area) of the field in the photographed image.
The normal color data includes color data before planting in a normal year and color data immediately before harvest in a normal year.
The color data before planting in a normal year is the color data on the day immediately before planting obtained from past data, and the wavelength components of the color data are as described above.
The color data immediately before harvest in a normal year is the color data on the day immediately before harvest obtained from past data, and the wavelength components of the color data are as described above.
The past data is, for example, image data of photographed fields for the past 10 years. If the number of years that have passed since the field was created is less than 10 years, this is the period from the time the field was created until the image is taken for predicting the amount of shipment.
Note that the normal color data is stored, for example, in the memory 36 shown in FIG. 2 or on the cloud.

Next, the amount of crop growth is calculated (step S26).
In step S26, it is calculated where the color of the field in the photographed image is positioned between the planting date and the day immediately before harvest, and the amount of growth is calculated based on the color of the field in the image.
When the amount of crop growth based on the color of the field in the image is _α1 , the amount of growth _α1 based on the color of the field in the image is expressed by the following formula when using the RGB wavelength components of the image data. .
α ₁ = ((Cr-Fr)/(Hr-Fr)+(Cg-Fg)/(Hg-Fg)+(Cb-Fb)/(Hb-Fb))/3

In the above formula, Cr, Cg, and Cb are the values of the wavelength components RGB of the color data on the specific day obtained in step S20.
Fr, Fg, and Fb are the values of the wavelength components RGB of the planting day color data for a normal year. The normal planting date color data is the color data of the planting date obtained from past data.
Moreover, the planting date in a normal year is the day when planting was performed in a normal year, and is the average day of planting dates determined from past data (that is, the planting date in a normal year).
Hr, Hg, and Hb are the values of wavelength components RGB of color data on the day immediately before harvest in a normal year. The color data on the day immediately before harvest in a normal year is the color data on the day immediately before harvest obtained from past data.
Further, the day immediately before harvest in a normal year is the day before the harvest date in a normal year, and is the average day immediately before harvest determined from past data (that is, the day before harvest in a normal year).

In step S26, the amount of growth is calculated from the number of days of growth.
When the amount of growth based on the number of days of growth is α ₂ , the amount of growth α ₂ based on the number of days of growth is expressed by the following formula using the planting date and the day immediately before harvest.
α ₂ = (D-Df)/(Dh-Df)
In the above formula, D is the specific date on which the image was taken. Df is the planting date in a normal year, and is the average planting date obtained from past data. Dh is the day immediately before harvest in a normal year, and is the average day immediately before harvest determined from past data.

Next, the growth rate of the crop is calculated (step S28).
When the growth rate of the crop is S, the growth rate S is expressed as S=α ₁ /α ₂ using α ₁ and α ₂ obtained in step S26 above.
When α ₁ >α ₂ , the period from the planting date to harvesting is short and the growth rate S is fast.
In addition, in order to calculate the growth rate of a specific area (specifically, a specific cultivation area), the growth rate is calculated at multiple locations in the area, and the average value thereof is calculated. For example, in the three cultivation areas 10a, 10b, and 10c shown in FIG. 1, the growth rate is calculated for each field 11a, 11b, and 11c. Then, the average value of the growth speeds of the plurality of fields 11a, 11b, and 11c is calculated for each of the three cultivation areas 10a, 10b, and 10c.
Furthermore, in order to calculate the growth rate of a specific vegetable, the growth rate is calculated over the entire production area where it is shipped during that period, and the average value thereof is calculated. For example, in a situation where the same vegetable is grown in the three cultivation regions 10a, 10b, and 10c shown in FIG. 1, the growth speed of each is calculated, and the average value thereof is calculated as the growth speed of the specific vegetable.

Next, as shown in FIG. 3, the yield of crops in the target area at the time of harvest is predicted based on the growth rate calculated this time and the past growth rate (step S14).
In step S14, the yield of crops in the target area is predicted, for example, using the correlation for each type of crop acquired in advance as described above. In this case, for example, the yields of the three cultivation areas 10a, 10b, and 10c of the production area 10 shown in FIG. 1 are predicted. That is, the yields of the fields 11a, 11b, and 11c shown in FIG. 1 are predicted.

Next, the shipping amount of a plurality of cultivation areas (that is, a plurality of target areas) is estimated (step S16).
In step S16, the yields of the three cultivation areas 10a, 10b, and 10c of the production area 10 shown in FIG. 1, which were predicted in step S14 described above, are used. Furthermore, if the period from the harvest date to the time of shipment to the market is known, it is possible to know on what date and how much of the crops harvested from the field will be shipped to the market. In step S16, for each of the fields 11a, 11b, and 11c in the three cultivation regions 10a, 10b, and 10c, the production area on the specific date is Estimate the amount of crops shipped to the market from 10.
In this way, based on the predicted harvest amount for each of the plurality of cultivation areas (plurality of target areas), the amount shipped to the market where crops are shipped from the plurality of cultivation areas is estimated. By estimating the amount shipped to the market, it is possible to adjust the amount shipped of the crops being produced to avoid oversupply or undersupply to the market. This makes it possible to suppress sudden rises and price collapses in crop prices.

<Application example of the present invention>
By applying the present invention, for example, in the case of lettuce, the amount shipped to the Tokyo market can be estimated if there is information on the amount shipped by prefecture shown below. Approximately 80% of shipments come from the following prefectures. Therefore, the target areas are preferably Nagano, Ibaraki, Shizuoka, Hyogo, Nagasaki, Gunma, Tochigi, Kagawa, Chiba, and Fukuoka.
Regarding other markets, it is possible to predict the shipment volume by investigating the cultivation areas (shipping sources) that account for more than half of the annual shipment volume, so select the cultivation areas that account for more than half of the shipment volume as target areas. It is preferable to do so.
The cultivation area to be investigated is preferably a production area that accounts for 60% or more of the annual shipment amount, and more preferably a cultivation area that accounts for 70% or more of the annual shipment amount.
Since the cultivation area of crops changes depending on the season, the cultivation area to be investigated may be changed for each season. In this case, cultivation areas covering the above-mentioned ratios are selected as target areas for each month.

The present invention is basically configured as described above. Although the shipment amount prediction method and shipment amount prediction system of the present invention have been described in detail above, the present invention is not limited to the above-described embodiments, and various improvements or changes can be made without departing from the gist of the present invention. Of course you can.

10 Production area 10a, 10b, 10c Cultivation area 11a, 11b, 11c, 12a Field 13 Crop 20 Shipping amount prediction system 22 Processing unit 24 Input unit 26 Display unit 30 Acquisition unit 31 Identification unit 32 Calculation unit 33 Prediction unit 34 Estimation unit 35 Display Control unit 36 Memory 37 Control unit 40, 41, 42, 43 Field

Claims

A step of identifying the harvest time of the crops in the target area based on an image taken from above of the target area where the crops are planted;
calculating the growth rate of the crop based on the harvest time;
predicting the harvest amount of the crop in the target area at the harvest time based on the growth rate calculated this time and the growth rate in the past;
A shipment amount prediction method comprising the step of estimating the shipment amount to a market where crops are shipped from a plurality of target areas based on the predicted harvest amount for each of the plurality of target areas.
The step of identifying the harvest time of the crop in the target area includes a current image of the target area where the crop is planted taken from above, and a current image of the target area where the crop is planted taken from above. The shipping amount prediction method according to claim 1, wherein a color distance from a past image is calculated, and the harvest time of the crop in the target area is specified based on the color distance.
Obtaining in advance the correlation between the yield of the crop in the target area at the time of past harvest and the growth rate,
3. The step of predicting the harvest amount of the crop in the target area at the harvest time uses the correlation to predict the harvest amount in the target area at the harvest time. Shipping volume forecasting method.
The shipping amount prediction method according to claim 1 or 2, wherein the crop is a leafy vegetable.
an identification unit that identifies the harvest time of the crops in the target area based on an image taken from above of the target area where the crops are planted;
a calculation unit that calculates the growth rate of the crop based on the harvest time;
a prediction unit that predicts the harvest amount of the crop in the target area at the harvest time based on the growth rate calculated this time and the growth rate in the past;
A shipment amount prediction system comprising: an estimation unit that estimates the shipment amount to a market to which crops are shipped from the plurality of target areas, based on the predicted harvest amount for each of the plurality of target areas.
The identification unit calculates a color distance between a current image of the target area where the crops are planted taken from above and a past image of the target area where the crops are planted taken from above. 6. The shipment amount prediction system according to claim 5, wherein a harvest time of the crop in the target area is specified based on the color distance.
Obtaining in advance the correlation between the yield of the crop in the target area at the time of past harvest and the growth rate,
The shipping amount prediction system according to claim 5 or 6, wherein the prediction unit uses the correlation to predict the harvest amount of the crop in the target area at the harvest time.
The shipment amount prediction system according to claim 5 or 6, wherein the crop is a leafy vegetable.