Harvest information prediction method, harvest information prediction program, and harvest information prediction system

Abstract

To accurately predict time required for harvest work in an entire cultivation range when fruit vegetable species are cultivated and harvested in each of a prescribed number of areas within the cultivation range.SOLUTION: An information processing device is configured to: select, as a representative area, from among specific areas included in a prescribed number of areas, an area in which a correlation between the way in which harvest work time fluctuates in the area and the way in which harvest work time fluctuates in an entire cultivation range satisfies prescribed conditions; estimate the number of fruits to be harvested from the representative area at a prescribed future timing based on the number of fruits and ripeness detected in the representative area; predict harvest work time required to harvest the estimated number of harvests in the representative area; and predict harvest work time for the entire cultivation range based on the predicted harvest work time for the representative area and a relation between the harvest work time for the representative area and the harvest work time for the entire cultivation range.SELECTED DRAWING: Figure 18

Description

[Technical Field]

[0001] This invention relates to a harvest information prediction method, a harvest information prediction program, and a harvest information prediction system. [Background technology]

[0002] Large-scale greenhouse horticulture facilities are equipped with environmental control systems that regulate temperature, humidity, CO2 concentration, and other factors. These systems control the cultivation environment to suit the crops, enabling efficient crop production.

[0003] Farms with a cultivated area of ​​approximately 1 hectare or more employ several dozen to a hundred people and operate in an organized manner. As farms become larger, the effective utilization of employees and labor-saving measures become increasingly important. In the cultivation of fruits and vegetables, harvesting accounts for more than 30% of total working hours, so if harvesting time can be predicted in advance, appropriate personnel allocation can be made, leading to increased efficiency. Furthermore, if harvest yield can be predicted, it can be used as a guideline for environmental control and shipment volume, and can also be an important factor in price negotiations with wholesalers.

[0004] Conventionally, there are known technologies for calculating the probability of fruit setting in tomatoes and bell peppers, and for outputting information on the number of harvestable fruits and information on fruit thinning based on the calculated probability of fruit setting (see, for example, Patent Document 1). In addition, there are known technologies for photographing fruits in cultivation facilities and counting mature fruits (see, for example, Patent Document 2). [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Japanese Patent Publication No. 2022-104601 [Patent Document 2] Patent No. 6673442 [Overview of the project] [Problems that the invention aims to solve]

[0006] In the case of Patent Document 1 mentioned above, the predicted number of fruits set is calculated from the probability of fruit set in a predetermined measurement range (1 plot), the harvestable week for the fruits predicted to set is identified based on environmental information, and the predicted total harvestable number for the entire greenhouse in each week is identified. For example, if it is predicted that x fruits can be harvested per 1 plot in a certain week, the total harvestable number for the entire greenhouse can be identified by multiplying this by the total number of plots in the greenhouse. However, since Patent Document 1 does not consider how to determine the measurement range (1 plot), there is a risk that the harvestable number cannot be identified accurately if there is variation in growth within the greenhouse.

[0007] The present invention aims to provide a harvest information prediction method, a harvest information prediction program, and a harvest information prediction system that can accurately predict the time required for harvesting work across an entire cultivation area when cultivating fruit vegetables in a predetermined number of areas within a cultivation area and harvesting the fruit. [Means for solving the problem]

[0008] The present invention provides a harvest information prediction method for predicting harvest information when cultivating fruit vegetables in a predetermined number of areas set within a cultivation range and harvesting the fruit, wherein the method involves selecting a representative area from among multiple areas included in the predetermined number of areas, where the correlation between the way in which the harvest work time in that area fluctuates and the way in which the harvest work time for the entire cultivation range fluctuates satisfies predetermined conditions, estimating the number of fruits to be harvested from the representative area at a predetermined time in the future based on the number of fruits and the degree of maturity detected in the representative area, predicting the harvest work time required to harvest the estimated number of fruits in the representative area, and predicting the harvest work time for the entire cultivation range based on the predicted harvest work time for the representative area and the relationship between the harvest work time for the representative area and the harvest work time for the entire cultivation range, with the processing performed by a computer. [Effects of the Invention]

[0009] The harvesting information prediction method, harvesting information prediction program, and harvesting information prediction system of the present invention have the effect of being able to accurately predict the time required for the harvesting operation of the entire cultivation range when cultivating fruit vegetables and harvesting fruits in each of a predetermined number of regions within the cultivation range.

Brief Description of the Drawings

[0010] [Figure 1] It is a diagram showing the configuration of a harvesting information prediction system according to an embodiment. [Figure 2] It is a diagram showing a state of a predetermined range (cultivation range) in a greenhouse as viewed from above. [Figure 3] FIG. 3(a) is a diagram showing the hardware configuration of an information processing device, and FIG. 3(b) is a diagram showing the hardware configuration of an operator terminal. [Figure 4] It is a functional block diagram of an operator terminal. [Figure 5] It is a diagram showing the hardware configuration of a fruit detection device. [Figure 6] It is a functional block diagram of an information processing device. [Figure 7] FIG. 7(a) is a diagram showing an example of a work DB, and FIG. 7(b) is a diagram showing an example of a work history DB. [Figure 8] FIG. 8(a) is a diagram showing an example of a specific area DB, and FIG. 8(b) is a diagram showing an example of a detection information DB. [Figure 9] It is a table showing the harvesting operation time in each specific area from the 1st week to the 10th week and the harvesting operation time in the entire area (the entire cultivation range). [Figure 10] FIG. 10(a) is a diagram showing the variation of the harvesting operation time per week in areas 65 and 66 and the entire cultivation range, and FIG. 10(b) is a diagram showing the relationship between the harvesting operation time in areas 65 and 66 and the harvesting operation time in the entire cultivation range. [Figure 11] FIG. 11(a) is a diagram showing the variation of the harvesting operation time per week in areas 117 and 118 and the entire cultivation range, and FIG. 11(b) is a diagram showing the relationship between the harvesting operation time in areas 117 and 118 and the harvesting operation time in the entire cultivation range. [Figure 12] It is a diagram showing the relationship between the harvesting work time in areas 35 and 36 and the harvesting work time for the entire cultivation area. [Figure 13] It is a diagram for explaining the change in the coefficient of determination for each area. [Figure 14] It is a graph showing the relationship between the number of fruits and the harvesting work time. [Figure 15] It is a graph showing the relationship between the total working time and the total harvest amount. [Figure 16] It is a flowchart showing the processing of the operator terminal. [Figure 17] It is a flowchart showing the processing of the information processing device. [Figure 18] FIG. 18(a) and FIG. 18(b) are diagrams showing an overview of the processing of one embodiment. [Figure 19] It is a flowchart showing the processing of the information processing device according to Modification 1. [Figure 20] FIGS. 20(a) to 20(c) are diagrams for explaining Modification 2.

Mode for Carrying Out the Invention

[0011] Hereinafter, an embodiment of the harvest information prediction system will be described in detail based on FIGS. 1 to 18. FIG. 1 schematically shows the configuration of a harvest information prediction system 100 according to an embodiment. The harvest information prediction system 100 of the present embodiment is a system for predicting the harvesting work time of fruit vegetables such as tomatoes and paprika, and predicting the harvest amount of fruit vegetables, in a large-scale facility (hereinafter referred to as a greenhouse) for cultivating fruit vegetables.

[0012] Figure 2 shows a view from above of a designated area (cultivation area) within a greenhouse. As shown in Figure 2, within the cultivation area, seedlings of fruit and vegetable crops are planted along the cultivation rows as shown in Figure 2. The area of ​​the greenhouse is, for example, 1 hectare or more, and the cultivation area in Figure 2 is assumed to be half the area of ​​the greenhouse. There are 130 cultivation rows within the cultivation area. There are aisles between each cultivation row, and workers harvest the fruit by moving back and forth along these aisles (see solid and dashed arrows). When harvesting, workers will move along the aisle between cultivation rows #1 and #2 to harvest rows #1 and #2. Also, workers will move along the aisle between cultivation rows #3 and #4 to harvest rows #3 and #4. Therefore, workers will not move along the aisle between cultivation rows #2 and #3 when harvesting. In other words, the worker moves along the path between cultivation rows #(2n-1) and #(2n), but does not move along the path between cultivation rows #(2n) and #(2n+1) (n=1, 2…).

[0013] In this embodiment, two adjacent planting rows are considered a single unit for harvesting and are referred to as a "region." Specifically, in this embodiment, planting rows #1 and #2 are collectively referred to as regions 1 and 2, planting rows #3 and #4 are collectively referred to as regions 3 and 4, and so on, with planting rows #129 and #130 being collectively referred to as regions 129 and 130.

[0014] As shown in Figure 1, the harvest information prediction system 100 comprises an information processing device 10 such as a PC (Personal Computer), a worker terminal 70 as a measuring device, and a fruit detection device 50. The information processing device 10, the worker terminal 70, and the fruit detection device 50 are connected to a network 80 such as the Internet or an LTE line.

[0015] (Information processing device 10) The information processing device 10 collects data from the worker terminal 70, including the harvesting time for each area by each worker and the amount of fruit harvested in each area. The information processing device 10 also collects data from the fruit detection device 50, including the number of fruits in each area and the maturity level of each fruit. Using the collected data, the information processing device 10 predicts the time required for harvesting the entire area (the entire cultivation area) and the total harvest amount for the entire cultivation area at a predetermined time in the future (for example, the following week).

[0016] Figure 3(a) shows the hardware configuration of the information processing device 10. As shown in Figure 3(a), the information processing device 10 includes a CPU 90, ROM 92, RAM 94, storage (SSD (Solid State Drive) or HDD (Hard Disk Drive)) 96, network interface 97, display unit 93, input unit 95, and portable storage medium drive 99, etc. Each of these components of the information processing device 10 is connected to a bus 98. In the information processing device 10, the CPU 90 executes a program (including a harvest information prediction program) stored in the ROM 92 or storage 96, or a program read from the portable storage medium 91 by the portable storage medium drive 99, thereby realizing the functions of each part shown in Figure 5. The functions of each part in Figure 5 may be realized by integrated circuits such as ASICs (Application Specific Integrated Circuits) or FPGAs (Field Programmable Gate Arrays). Details of Figure 5 will be described later.

[0017] (Worker terminal 70) The worker terminal 70 is a terminal carried or worn on the worker's body and has the hardware configuration shown in Figure 3(b). As shown in Figure 3(b), the worker terminal 70 includes a CPU 190, ROM 192, RAM 194, storage (SSD or HDD) 196, network interface 197, display unit 193, input unit 195, sensor 189, and portable storage medium drive 199, etc. In the worker terminal 70, the CPU 190 executes programs stored in the ROM 192 or storage 196, or programs read from the portable storage medium by the portable storage medium drive 199, thereby realizing the functions of each part shown in Figure 4.

[0018] As shown in Figure 4, the worker terminal 70 functions as a region information reading unit 170, a timing unit 172, a harvest amount acquisition unit 174, and a work information transmission unit 176 when the CPU 190 executes a program. Figure 4 also shows the work DB 180 stored in the storage 196, etc.

[0019] As shown in Figure 2, the area information reading unit 170 reads information from the identifiers 71 and 72 when the sensor 189 recognizes the entrance identifier 71 and the exit identifier 72 provided in the passages of each area (which can also be called the entrance and exit parts of each area). In this embodiment, the identifiers include an area ID indicating which area they are installed in, and information indicating whether they are entrances or exits.

[0020] When the area information reading unit 170 reads an entry identifier, the timing unit 172 obtains the area ID and starts timing the harvesting work time for that area. Similarly, when the area information reading unit 170 reads an exit identifier, the timing unit 172 obtains the area ID and stops timing the harvesting work time for that area. The timing unit 172 stores the area ID and the harvesting work time for that area in the work database 180 shown in Figure 7(a). This work database 180 is created for each worker and each week, and stores information on the harvesting work time and harvest yield for each area.

[0021] The harvest yield acquisition unit 174 receives the harvest yield (weight) via the input unit 195 after the worker reads the identifier of the exit of a certain area. The harvest yield acquisition unit 174 may measure immediately after or during harvesting by placing a platform scale or the like at the exit of the area or on a work cart, or it may measure during sorting or processing work. Alternatively, the harvest yield acquisition unit 174 may maintain a relational expression showing the relationship between the number of containers and the weight, and estimate the harvest yield (weight) by inputting the number of containers into this relational expression.

[0022] The harvest yield acquisition unit 174 stores the measured or estimated harvest yield (weight) in the work DB 180 shown in Figure 7(a).

[0023] The work information transmission unit 176 transmits information stored in the work DB 180 to the information processing device 10 at predetermined intervals. The information processing device 10 collects information transmitted from each worker's worker terminal 70.

[0024] (Fruit detection device 50) Figure 5 shows the hardware configuration of the fruit detection device 50. The fruit detection device 50, for example, moves around within an area at night, taking images and detecting the number and ripeness of fruits present in the area. As shown in Figure 5, the fruit detection device 50 comprises a travel device 150, a position detection device 152, a camera 154, a lighting device 156, a control device 160, and a communication device 158.

[0025] The traveling device 150 has wheels and a motor to drive the wheels, and travels at a constant speed along rails installed between the cultivation rows under the direction of the control device 160. The components of the fruit detection device 50 other than the traveling device 150 are mounted on the main body attached to the traveling device 150 and move together with the traveling device 150.

[0026] The position detection device 152 is a device that detects the position of the traveling device 150. The control device 160 controls the timing of the camera 154 to take pictures based on the detection result from the position detection device 152.

[0027] Camera 154 photographs the plants of fruit and vegetables and outputs the captured images to the control device 160. Camera 154 photographs all plants present in the area while the fruit detection device 50 is moving within the area. Illumination device 156 illuminates the shooting range of camera 154.

[0028] The control device 160 controls the driving of the traveling device 150 and detects the number of fruits present in the area and the ripeness of each fruit from the images captured by the camera 154. In this process, the control device 160 uses object detection technology based on deep learning. Furthermore, when detecting ripeness, the degree of coloration of the fruit is used. The detection of the number of fruits and ripeness from the images may be performed in real time while the vehicle is moving, or it may be performed on images captured after the fact.

[0029] When the number of fruits and their ripeness in a given area are detected, the communication device 158 transmits the detection results to the information processing device 10 under the instruction of the control device 160.

[0030] Furthermore, if the fruiting vegetables are crops such as bell peppers where the fruit-bearing position is high, the fruit detection device 50 may be equipped with a mechanism to move the camera 154 and lighting device 156 vertically in order to photograph the entire plant. Alternatively, the fruit detection device 50 may be mounted on a separate, independent traveling device that has a mechanism for moving it vertically. The traveling device 150 does not have to move along rails, but may also travel on the ground.

[0031] (Regarding the functions of the information processing device 10) Figure 6 shows a functional block diagram of the information processing device 10. As shown in Figure 6, the information processing device 10 functions as a work information receiving unit 20, a detection instruction unit 22, a detection information receiving unit 24, a representative area selection unit 26 as a selection unit, a prediction unit 28, and an output unit 30 when the CPU 90 executes a program. Figure 6 also illustrates the work history DB 40, specific area DB 42, and detection information DB 44 stored in the storage 96, etc.

[0032] The work information receiving unit 20 receives information transmitted from the work information transmitting unit 176 of the worker terminal 70 and stores it in the work history DB 40. Here, the work history DB 40 has a data structure as shown in Figure 7(b). The work history DB 40 collects the data stored in the work DB 180 of each worker terminal 70 as shown in Figure 7(a). In other words, the work history DB 40 stores information on harvest work time and harvest yield for each area each week. The work history DB 40 also stores the total harvest work time for each area and the total harvest yield for each area, that is, the total harvest work time and harvest yield for the entire cultivation area shown in Figure 2.

[0033] The detection instruction unit 22 checks the information of the specific area stored in the specific area DB 42 and instructs the fruit detection device 50 to detect the number of fruits and their ripeness in the specific area. In other words, in this embodiment, the number of fruits and their ripeness are not detected in all areas, but only in predetermined specific areas. As shown in Figure 8(a), the specific area DB 42 stores identification information for a predetermined number of areas (for example, three) as specific areas. The specific areas may be multiple areas randomly selected from all areas, or multiple areas arranged arbitrarily.

[0034] The detection information receiving unit 24 receives the number of fruits and their maturity level in a specific area detected by the fruit detection device 50 and stores them in the detection information DB 44. The detection information DB 44 has a data structure like that shown in Figure 8(b), for example. Specifically, the detection information DB 44 stores, linked to the area ID of the specific area, the number of fruits with a maturity level of 100% to 90%, the number of fruits with a maturity level of 90% to 80%, the number of fruits with a maturity level of 80% to 70%, and so on.

[0035] The representative area selection unit 26 refers to the work history DB 40 and selects a representative area (an area that uses the detection results of the fruit detection device 50) from among the specific areas stored in the specific area DB 42. The representative area is the area used to predict the harvesting work time and harvest yield for the entire area (the entire cultivation area). As an example, Figure 9 shows the harvesting work time for each specific area from week 1 to week 10, and the harvesting work time for the entire area (the entire cultivation area). The representative area selection unit 26 checks the correlation between how the harvesting work time changes in each specific area shown by the thick solid line frame and how the harvesting work time changes in the entire area shown by the thick dashed line frame.

[0036] In the specific areas 65 and 66, harvesting time fluctuates as shown by the solid line in Figure 10(a). Harvesting time for the entire area (the entire cultivation range) also fluctuates as shown by the dashed line in Figure 10(a). Figure 10(b) shows a graph plotting weekly values ​​with harvesting time for specific areas 65 and 66 on the horizontal axis and harvesting time for the entire cultivation range on the vertical axis. The thick solid line in Figure 10(b) is an approximate straight line obtained using the least squares method, etc. As shown in Figure 10(b), the coefficient of determination R 2 Since the value is 0.57, it can be seen that there is not much correlation between the harvesting time in specific areas 65 and 66 and the harvesting time for the entire cultivation area.

[0037] In the specific areas 117 and 118, the harvesting time fluctuates as shown by the solid line in Figure 11(a). The harvesting time for the entire area (the entire cultivation range) also fluctuates as shown by the dashed line in Figure 11(a). Figure 11(b) shows a graph plotting weekly values ​​with the harvesting time for specific areas 117 and 118 on the horizontal axis and the harvesting time for the entire cultivation range on the vertical axis. The thick solid line in Figure 11(b) is an approximate straight line obtained using the least squares method, etc. As shown in Figure 11(b), the coefficient of determination R 2 Since the value is 0.94, it can be seen that there is a high correlation between the harvesting time in specific areas 117 and 118 and the harvesting time for the entire cultivation area.

[0038] FIG. 12 shows a graph in which the horizontal axis represents the harvesting operation time of specific regions 35 and 36, the vertical axis represents the harvesting operation time of the entire cultivation range, and the values for each week are plotted. The thick solid line in FIG. 12 is an approximate straight line obtained using the least squares method or the like. As shown in FIG. 12, since the coefficient of determination R 2 is 0.7179, it can be seen that there is a high correlation between the harvesting operation time of specific regions 117 and 118 and the harvesting operation time of the entire cultivation range.

[0039] In the present embodiment, the representative region selection unit 26 selects a representative region from among the specific regions based on the coefficient of determination R 2 . For example, if a selection condition is set to select two specific regions with a large coefficient of determination R 2 , the representative region selection unit 26 selects specific regions 117 and 118 and specific regions 3 and 36 as representative regions. Also, for example, if a selection condition is set to select those specific regions for which the coefficient of determination R 2 is greater than a threshold value (for example, 0.8), the representative region selection unit 26 selects only specific regions 117 and 118 as representative regions.

[0040] FIG. 13 shows the coefficient of determination R 2 for each region. In FIG. 13, the coefficient of determination R 2 when a graph as shown in FIG. 12 is created using the harvesting operation time from the 11th week to the 3rd week, the coefficient of determination R 2 when a graph as shown in FIG. 12 is created using the harvesting operation time from the 1st week to the 4th week, and so on are shown. As shown in FIG. 13, the coefficient of determination R 2 changes as the week progresses. Therefore, when the representative region selection unit 26 selects a representative region, it uses the latest coefficient of determination R 2 .

[0041] Returning to Figure 6, the prediction unit 28 uses the number of fruits and maturity levels of the representative area stored in the detection information DB 44 to predict the number of fruits to be harvested in the representative area at a predetermined time in the future (for example, the following week). For example, if only areas 117 and 118 are selected as the representative area, the prediction unit 28 refers to the row with area ID = "117,118" in the detection information DB 44 in Figure 8 to predict the number of fruits that can be harvested the following week. If it is known that the fruits that can be harvested the following week are those with a detected maturity level of 80% or higher, the prediction unit 28 predicts from Figure 8 that the number of fruits that can be harvested the following week will be (Z1 + Z2). Note that the method for predicting the number of fruits to be harvested the following week is not limited to the calculation method described above; other calculation methods may also be used.

[0042] Furthermore, the prediction unit 28 predicts the harvesting time for the following week in representative regions 117 and 118 based on the number of fruits harvested (Z1+Z2) in those regions. At this time, the prediction unit 28 predicts the harvesting time for the following week in representative regions 117 and 118 by inputting the number of fruits harvested (Z1+Z2) into a graph (equation) showing the relationship between the number of fruits and the harvesting time, as shown in Figure 14. Note that each point in Figure 14 is a plot on a coordinate system of the relationship between the measured number of fruits harvested in an arbitrary region and the measured time required for harvesting in that region (harvesting time), and the thick solid line graph in Figure 14 is an approximate straight line calculated from each point.

[0043] Furthermore, the prediction unit 28 predicts the estimated harvest time (total work time) for the entire cultivation area for the following week based on the estimated harvest work time for the following week in representative areas 117 and 118. In this case, the prediction unit 28 predicts the estimated harvest time (total work time) for the entire cultivation area for the following week by inputting the estimated harvest work time for the following week in representative areas 117 and 118 into the graph (equation) shown in Figure 11(b).

[0044] Furthermore, the prediction unit 28 predicts the yield (total yield (weight)) for the following week across the entire cultivation area, based on the predicted harvest time (total working time) for the following week across the entire cultivation area. At this time, the prediction unit 28 predicts the total yield for the following week by inputting the predicted total working time into a graph (equation) showing the relationship between total working time and total yield, as shown in Figure 15. Note that each point in Figure 15 is a plot on a coordinate system representing the relationship between the measured total working time for each week and the measured total yield for each week, and the thick solid line graph in Figure 15 is an approximate straight line calculated from each point.

[0045] The output unit 30 displays the total working hours for the following week and the total harvest amount for the following week, as predicted by the prediction unit 28, on the display unit 93.

[0046] Furthermore, if multiple representative areas are selected, the output unit 30 outputs, for example, the average of the total work time predicted from each representative area as the total work time, and the average of the total yield predicted from each representative area as the total yield. However, the output unit 30 is not limited to this, and outputs the coefficient of determination R for the total work time predicted from each representative area. 2 Alternatively, the weighted average (weighted mean) can be calculated by multiplying it by the corresponding weight coefficients, and this value can be output as the total work time. Alternatively, the coefficient of determination R can be applied to the total yield predicted from each representative region. 2 Alternatively, you can multiply by the corresponding weighting coefficients to calculate a weighted average and output that value as the total yield.

[0047] (Regarding the processing on worker terminal 70) Next, the processing of the worker terminal 70 will be explained according to the flowchart in Figure 16. The worker carries the worker terminal 70 and performs harvesting work in each area. At the start of work in each area, the worker reads the identifier 71 of the entrance to that area using the sensor 189 of the worker terminal 70, and at the end of work in each area, the worker reads the identifier 72 of the exit to that area using the sensor 189. In addition, after finishing harvesting work in one area and before moving on to harvesting work in the next area, the worker inputs the weight of the fruit harvested in the area where harvesting work has been completed via the input unit 195 of the worker terminal 70.

[0048] When the process shown in Figure 16 begins, first, in step S10, the region information reading unit 170 waits until it has read the identifier 71 at the entrance of each region. Once the operator has read the identifier 71 using the sensor 189, the process proceeds to step S12.

[0049] When the process moves to step S12, the timing unit 172 starts timing.

[0050] Next, in step S14, the area information reading unit 170 waits until it reads the identifier 72 of the area exit. Once the operator reads the identifier 72 using the sensor 189, the process proceeds to step S16.

[0051] When the process moves to step S16, the timing unit 172 terminates the timing.

[0052] Next, in step S18, the timing unit 172 stores the measured harvesting time in the work DB 180, linked to the area IDs included in identifiers 71 and 72 (see Figure 7(a)).

[0053] Next, in step S20, the harvest amount acquisition unit 174 waits until the harvest amount (weight) is input. Once the operator inputs the weight of the fruit harvested within the area, the process proceeds to step S22.

[0054] When the process moves to step S22, the harvest yield acquisition unit 174 stores the harvest yield (weight) in the work DB 180, linked to the area ID.

[0055] Next, in step S24, the work information transmission unit 176 determines whether or not it is time to transmit work information. If the determination in step S24 is negative, the process returns to step S10. Here, the transmission timing is assumed to be, for example, once a week or once a day.

[0056] On the other hand, if the judgment in step S24 is affirmed, the process proceeds to step S26, where the work information transmission unit 176 transmits work information to the information processing device 10. That is, the work information transmission unit 176 transmits to the information processing device 10 any information that has been newly added to the work DB 180 since the previous transmission timing. After that, the process returns to step S10.

[0057] (Processing by the information processing device 10) Next, the processing of the information processing device 10 will be explained according to the flowchart in Figure 17. Note that the work information receiving unit 20 in Figure 6 receives information from the work DB 180 transmitted from the worker terminal 70 and stores it in the work history DB 40 (Figure 7(b)), separately from the processing flow in Figure 17.

[0058] When the process shown in Figure 17 begins, first, in step S50, the detection instruction unit 22 determines whether or not it is time for the fruit detection device 50 to detect fruit. The detection timing is, for example, a predetermined time (nighttime) when the worker is not performing harvesting work. If the detection instruction unit 22 determines that it is time for detection, it proceeds to step S52.

[0059] When the process moves to step S52, the detection instruction unit 22 instructs the fruit detection device 50 to detect the number and ripeness of fruits in a specific area defined in the specific area DB42. In response to this instruction, the fruit detection device 50 takes images while driving through the specific area and analyzes the images to detect the number and ripeness of fruits within the specific area.

[0060] Next, in step S54, the detection information receiving unit 24 receives the detection results for the number of fruits and their maturity and stores them in the detection information DB 44.

[0061] Next, in step S55, the representative area selection unit 26 determines whether or not a prediction request has been entered by the operator. When an operator wants to check the total harvest time or total harvest amount at a predetermined time in the future (such as the following week), they input this information via the input unit 95. If this input is received and the determination in step S55 is affirmative, the representative area selection unit 26 proceeds to step S56. If the determination in step S55 is negative, it returns to step S50. If the determination in step S50 is also negative, it proceeds to step S55. In other words, steps S50 (negative) and S55 (negative) are repeatedly executed until the detection timing is reached or until a prediction request is received.

[0062] When a prediction request is input from the worker and the process moves to step S56, the representative area selection unit 26 executes a process to select a representative area from the specified areas. Specifically, the representative area selection unit 26 calculates the correlation between the variation in harvesting work time for each specified area and the variation in harvesting work time for the entire area, based on the data stored in the work history DB 40. Then, it selects a specified area whose correlation satisfies predetermined conditions as the representative area.

[0063] Next, in step S58, the prediction unit 28 identifies one unspecified representative region.

[0064] Next, in step S60, the prediction unit 28 reads the number of fruits and their maturity in the identified representative area from the detection information DB 44, and predicts the number of fruits to be harvested in the representative area at a predetermined time in the future based on the read information.

[0065] Next, in step S62, the prediction unit 28 predicts the harvesting time for a representative area at a predetermined timing based on the number of harvests for the representative area predicted in step S60. At this time, the prediction unit 28 predicts the harvesting time for the representative area by inputting the number of harvests (predicted value) for the representative area into the formula shown in Figure 14.

[0066] Next, in step S64, the prediction unit 28 predicts the total work time for all areas at a predetermined timing based on the harvest work time for the representative area predicted in step S62. At this time, the prediction unit 28 predicts the harvest work time for all areas by inputting the harvest work time (predicted value) for the representative area into the equation shown in Figure 11(b).

[0067] Next, in step S66, the prediction unit 28 predicts the total harvest yield (weight) at a predetermined timing based on the total working time for the entire area predicted in step S64. At this time, the prediction unit 28 predicts the total harvest yield for the entire area by inputting the predicted harvesting time for the entire area into the formula shown in Figure 15.

[0068] Next, in step S68, the prediction unit 28 determines whether all representative regions have been identified. If the determination in step S68 is negative, the process returns to step S58 and the processing and determination in steps S58 to S68 are repeated. On the other hand, if the determination in step S68 is positive, the process proceeds to step S70.

[0069] When the system moves to step S70, the output unit 30 outputs the prediction results. In this case, if only one representative area was selected in step S56, the output unit 30 outputs the prediction results (total work time and total yield) from steps S64 and S66. On the other hand, if two or more representative areas were selected in step S56, the output unit 30 outputs the average value and weighted average value of total work time, the average value and weighted average value of total yield, etc., obtained from the information of each representative area.

[0070] After the process in step S70 is completed, the process returns to step S50 and the process described above is repeated.

[0071] In the explanation of Figure 17 described above, the case in which the information processing device 10 (detection instruction unit 22) instructs the fruit detection device 50 to detect the number of fruits and their ripeness was described, but this is not the only case. The fruit detection device 50 may, for example, start detecting the number of fruits and their ripeness in response to input from an operator. Alternatively, the fruit detection device 50 may start detecting the number of fruits and their ripeness at a preset timing.

[0072] Furthermore, while the explanation described the case in which the information processing device 10 executes the processing from step S56 onwards when the operator inputs a prediction request (S55: affirmative), it is not limited to this case. For example, the operator can also pre-determine the prediction interval and prediction conditions (when the future timing to be predicted should be). In this case, the information processing device 10 should execute the processing from step S55 onwards based on the pre-determined prediction interval and prediction conditions.

[0073] As described in detail above, according to this embodiment, the representative area selection unit 26 selects a representative area from among specific areas set within the cultivation area in which the correlation between the way the harvesting work time of the area fluctuates and the way the harvesting work time of the entire cultivation area fluctuates satisfies predetermined conditions (S56). The prediction unit 28 then predicts the number of fruits to be harvested from the representative area at a predetermined timing in the future, based on the number of fruits and maturity detected in the representative area (S60). Furthermore, the prediction unit 28 predicts the harvesting work time required to harvest the number of fruits (predicted value) in the representative area (S62), and predicts the harvesting work time of the entire cultivation area based on the predicted harvesting work time of the representative area (S64). Thus, in this embodiment, since an area correlated with the entire cultivation area is selected as the representative area, the harvesting work time of the entire cultivation area can be predicted with high accuracy by predicting the number of fruits to be harvested from the number of fruits and maturity detected in this representative area.

[0074] Furthermore, in this embodiment, the prediction unit 28 predicts the total harvest yield for the entire cultivation area based on the predicted harvest time for the entire cultivation area, using the relationship between harvesting time and harvest yield (Figure 15). In this case as well, since a region correlated with the entire cultivation area is selected as a representative region, the total harvest yield for the entire cultivation area can be predicted with high accuracy by predicting the harvest yield for the entire cultivation area based on the number of fruits and maturity levels detected in this representative region.

[0075] Thus, in this embodiment, even without detecting the number of fruits or their maturity across the entire cultivation area, the total harvesting time (total work time) and harvest yield (total harvest yield) for the entire cultivation area can be predicted with high accuracy, enabling efficient prediction of total work time and total harvest yield. Furthermore, by accurately predicting total work time and total harvest yield, it is possible to appropriately allocate personnel at predetermined timings in the future. In addition, the prediction results can be used as a guideline for environmental control and shipment volume, and can be used as important decision-making material in price negotiations with wholesalers.

[0076] Furthermore, in this embodiment, as shown in Figure 18(a), a specific region is predetermined for detecting the number of fruits and their maturity, and as shown in Figure 18(b), the coefficient of determination R is determined from within this specific region. 2 A representative area is selected based on this. In this case, the fruit detection device 50 only needs to detect fruit in a fixed, specific area. That is, since the fruit detection device 50 does not travel outside the specific area, the number of places where rails for the fruit detection device 50 to travel can be reduced. Furthermore, even if the fruit detection device 50 is of a type that is always installed and operated only in a specific area, the results measured by that device can be used for prediction. Since a representative area is selected from the fixed area and used for prediction, the accuracy of predictions for total working time and total harvest can be improved compared to cases where all fixed areas are used for prediction or where areas to be used for prediction are randomly selected from the fixed area.

[0077] In the above embodiment, the case in which the worker terminal 70 reads information contained in the identifier when timing the harvesting work is described, but it is not limited to this. For example, the worker terminal 70 may read information from a storage medium such as an IC chip provided in each area, or it may read information by taking a picture of a marker such as a two-dimensional code with the camera of a terminal such as a smartphone, or it may read information contained in the identifier using a harvesting work cart (a mobile cart on which the worker rides or a cart for transporting the harvested goods) equipped with sensors, etc. In addition, the worker may manually input the start and end times of the harvesting work to the worker terminal 70. Furthermore, cameras and sensors that take an overhead view of the entire area or the entrance and exit of the area may be installed to recognize the worker's movements and time the harvesting work in each area.

[0078] (Variation 1) In the above embodiment, a specific region is predetermined, and the coefficient of determination R described above is selected from within that specific region. 2 While we have described the case in which a representative area is selected based on the above, it is not limited to this. For example, the information processing device 10 may perform the processing shown in Figure 19. In Figure 19, processing that differs from the processing in Figure 17 is indicated by a thick border. Also, in Figure 19, step S56 of Figure 17 is omitted.

[0079] In the process shown in Figure 19, unlike the embodiment described above, a specific region is not predetermined. Instead, each time the fruit detection device 50 detects a fruit (each time the judgment in step S50 is affirmed), the coefficient of determination R is determined from the entire region. 2 A representative region is selected based on this (see S51 in Figure 19). In this case, the coefficient of determination R 2 A predetermined number of representative regions will be determined in descending order of the coefficient of determination R. 2 The entire region where the threshold is greater may be designated as the representative region.

[0080] The fruit detection device 50 detects the number of fruits and their ripeness in a representative area (S52' in Figure 19). Then, if a prediction request is received from the operator, the total working time and total harvest amount are predicted using the data (number of fruits, ripeness, and harvesting time) of the representative area selected in step S51, as shown in step S58 and beyond.

[0081] As described above, according to this modified example 1, R 2 A region with a high probability can be selected as the representative region each time. Therefore, the total working time and total yield can be predicted with high accuracy.

[0082] (Modification 2) In the above embodiment, we have described a case where the specific region is fixed and does not change, but this is not the only case. For example, if a specific region is initially defined as shown in Figure 20(a), and a representative region is selected as shown in Figure 20(b), the selected region will remain as part of the specific region, but the region not selected as the representative region may be excluded from the specific region. In this case, as shown in Figure 20(c), other regions will be added as part of the specific region. The other regions to be added to the specific region are R 2 Based on this, for example, R 2 This can be the largest region. In this way, a region with a high correlation to the entire cultivation area can always be used as a specific region. However, this is not the only option; other regions to be added to the specific region may be randomly selected from all regions.

[0083] The above processing functions can be implemented by a computer. In this case, a program describing the processing content of the functions that the processing unit should have is provided. By executing this program on a computer, the above processing functions are implemented on the computer. The program describing the processing content can be recorded on a storage medium that can be read by a computer (except for carrier waves).

[0084] When distributing a program, it may be sold in the form of a portable storage medium such as a DVD (Digital Versatile Disc) or CD-ROM (Compact Disc Read Only Memory) on which the program is recorded. Alternatively, the program can be stored in the storage device of a server computer and transferred from the server computer to other computers via a network.

[0085] A computer executing a program stores programs, for example, those recorded on a portable storage medium or transferred from a server computer, in its own memory. The computer then reads the program from its memory and executes the processing according to the program. Alternatively, the computer can directly read the program from the portable storage medium and execute the processing according to that program. Furthermore, the computer can sequentially execute the processing according to the programs received as they are transferred from the server computer.

[0086] The embodiments described above are preferred examples of the present invention. However, the invention is not limited thereto, and various modifications are possible without departing from the spirit of the invention. [Explanation of symbols]

[0087] 10 Information Processing Devices 26 Representative area selection department (selection department) 28 Prediction Section 50 Fruit detection device 70. Worker terminal (measuring device) 100 Harvest Information Prediction System

Claims

1. A harvest information prediction method for predicting information related to harvesting when cultivating fruit vegetables in each of a predetermined number of areas set within a cultivation range and harvesting the fruit, Among the multiple regions included in the predetermined number of regions, a region is selected as a representative region in which the correlation between the way harvesting work time varies in that region and the way harvesting work time varies in the entire cultivation area satisfies predetermined conditions. Based on the number of fruits and their maturity detected in the aforementioned representative region, the number of fruits to be harvested from the representative region at a predetermined time in the future is estimated. The harvesting time required to harvest the estimated yield in the representative area is predicted, and the harvesting time for the entire cultivation area is predicted based on the predicted harvesting time for the representative area and the relationship between the harvesting time for the representative area and the harvesting time for the entire cultivation area. A method for predicting harvest information, characterized in that the processing is performed by a computer.

2. The harvest information prediction method according to claim 1, characterized in that a computer performs a process of predicting the total harvest amount in the entire cultivation area from the predicted total harvest time in the entire cultivation area, based on the relationship between the harvest work time and the harvest amount.

3. The aforementioned multiple regions are fewer than the predetermined number of regions. The harvest information prediction method according to claim 1 or 2, characterized in that the number of fruits and the degree of maturity are detected in the multiple regions, and the number of fruits and the degree of maturity are not detected in regions other than the multiple regions.

4. The aforementioned multiple regions are all of the predetermined number of regions, The harvest information prediction method according to claim 1 or 2, characterized in that the number of fruits and the degree of maturity are detected in the representative region among the multiple regions in which the correlation satisfies predetermined conditions, and the number of fruits and the degree of maturity are not detected in regions other than the representative region.

5. The aforementioned multiple regions are fewer than the predetermined number of regions. The number of fruits and the degree of maturity are detected in the aforementioned multiple regions, while the number of fruits and the degree of maturity are not detected in regions other than the aforementioned multiple regions. In the selection process described above, the region that was not selected as the representative region from among the multiple regions is excluded from the multiple regions, and the other regions are added to the multiple regions. The harvest information prediction method according to claim 1 or 2, characterized by the above.

6. A harvest information prediction program that predicts information regarding harvesting when cultivating fruit vegetables in each of a predetermined number of areas set within a cultivation range and harvesting the fruit, Among the multiple regions included in the predetermined number of regions, a region is selected as a representative region in which the correlation between the way harvesting work time varies in that region and the way harvesting work time varies in the entire cultivation area satisfies predetermined conditions. Based on the number of fruits and their maturity detected in the aforementioned representative region, the number of fruits to be harvested from the representative region at a predetermined time in the future is estimated. The harvesting time required to harvest the estimated yield in the representative area is predicted, and the harvesting time for the entire cultivation area is predicted based on the predicted harvesting time for the representative area and the relationship between the harvesting time for the representative area and the harvesting time for the entire cultivation area. A harvest information prediction program characterized by having a computer perform the processing.

7. The harvest information prediction program according to claim 6, characterized in that it causes a computer to perform a process of predicting the total harvest amount in the entire cultivation area from the predicted total harvest time in the entire cultivation area, based on the relationship between the harvest work time and the harvest amount.

8. A harvest information prediction system that predicts information regarding harvesting when cultivating fruit vegetables in each of a predetermined number of areas set within a cultivation range and harvesting the fruit, A fruit detection device that travels through the aforementioned region and detects the number and ripeness of fruits in that region, A measuring device for measuring the harvesting time in each of the predetermined number of areas, The system comprises an information processing device that predicts harvest information based on the detection results of the fruit detection device and the measurement results of the measuring device, The aforementioned information processing device is A selection unit selects, from among multiple regions included in the predetermined number of regions, a region in which the correlation between the way harvesting work time varies in that region and the way harvesting work time varies in the entire cultivation area satisfies predetermined conditions as a representative region. The system includes a prediction unit that estimates the number of fruits to be harvested from the representative area at a predetermined future timing based on the number of fruits and their maturity detected in the representative area, predicts the harvesting time required to harvest the estimated number of fruits in the representative area, and predicts the harvesting time for the entire cultivation area based on the predicted harvesting time for the representative area and the relationship between the harvesting time for the representative area and the harvesting time for the entire cultivation area. A harvest information prediction system characterized by the following features.

9. The harvest information prediction system according to claim 8, characterized in that the prediction unit predicts the harvest yield for the entire cultivation area from the predicted harvest time for the entire cultivation area based on the relationship between the harvest work time and the harvest yield.

Images