Determination device, determination method, and program

Abstract

To determine fruits which should be thinned on the basis of a clear and reasonable determination reference for a manager and a cultivator of an orchard, while considering profitability for the fruit, in cultivation process of fruit cultivation.SOLUTION: A determination device comprises: an acquisition unit which acquires an image in which a plurality of fruits of crops are captured; a prediction unit which predicts sizes and sugar contents of the fruits left after the fruits are thinned at a future time point, in each pattern in which one or more fruits have been thinned from the plurality of fruits, on the basis of the image; and a determination unit which calculates an index value related to revenue of the fruits left after the fruits are thinned, on the basis of the sizes and the sugar contents of the fruits left after the fruits are thinned at the predicted future time point, and determines the fruits to be thinned among the plurality of fruits, on the basis of the calculated index value.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a determination device, a determination method, and a program. [Background technology]

[0002] Conventionally, there is known a technique for suggesting which fruits should be thinned in the process of fruit cultivation in order to improve marketability such as appearance and taste. For example, Patent Document 1 describes a technique for suggesting which flowers should be thinned by comparing an image captured to include the central flower and side flowers of an apple tree with a reference image captured in the past to include the central flower and side flowers. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2022-111047 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, the conventional techniques do not suggest which fruits should be thinned in consideration of the profitability of the cultivated fruits. Furthermore, in the conventional techniques, the fruits to be thinned and the amount of thinning are often determined by the sole judgment of an experienced fruit grower, and it is sometimes impossible for orchard managers and growers to determine which fruits should be thinned and the amount of thinning based on clear and rational judgment criteria.

[0005] The present invention has been made in consideration of the above circumstances, and one of its objectives is to provide a judgment device, judgment method, and program that can determine which fruits should be thinned during the fruit cultivation process based on clear and rational judgment criteria that are convenient for orchard managers and growers, while taking into account the profitability of the fruits. [Means for solving the problem]

[0006] A judgment device according to one embodiment of the present invention comprises an acquisition unit that acquires images of multiple fruits of a crop; a prediction unit that predicts the size and sugar content of the fruit remaining after thinning at a future time point for each pattern in which one or more fruits are thinned from the multiple fruits based on the images; and a judgment unit that calculates an index value related to the profit of the fruit remaining after thinning based on the predicted size and sugar content of the fruit remaining after thinning at the future time point, and judges which of the multiple fruits to thin based on the calculated index value.

[0007] The acquisition unit may acquire multiple images of the multiple fruits taken in time series, and the prediction unit may predict the size of the fruits remaining after the fruit thinning at the future point in time based on the growth rate of each of the multiple fruits captured in the multiple images.

[0008] The prediction unit may predict the sugar content of the fruits remaining after the fruit thinning at the future time point based on the fruit setting positions of each of the plurality of fruits.

[0009] The prediction unit may predict that the sugar content of the fruit remaining after the fruit thinning at the future time will be higher the higher the fruit-setting position of each of the plurality of fruits.

[0010] The prediction unit may predict the sugar content of the fruits remaining after the fruit thinning at the future time based on the fruit setting density of the plurality of fruits.

[0011] The prediction unit may predict that the sugar content of the fruit remaining after the fruit thinning at the future time will be higher as the fruit-bearing density of the plurality of fruits is lower.

[0012] The judgment unit may calculate an expected profit from the fruits remaining after the fruit thinning as the index value, and determine which fruits to thin from among the plurality of fruits so as to maximize the calculated expected profit.

[0013] The crop may be an apple.

[0014] In one embodiment of the present invention, a determination method involves a computer acquiring images of multiple fruits of a crop, predicting, based on the images, the size and sugar content of the fruit remaining after thinning at a future time point for each pattern in which one or more fruits are thinned from the multiple fruits, calculating an index value related to the profitability of the fruit remaining after thinning based on the predicted size and sugar content of the fruit remaining after thinning at the future time point, and determining which of the multiple fruits to thin based on the calculated index value.

[0015] A program according to one embodiment of the present invention causes a computer to acquire images of multiple fruits of a crop, predict based on the images the size and sugar content of the fruit remaining after thinning at a future time point for each pattern in which one or more fruits are thinned from the multiple fruits, calculate an index value related to the profitability of the fruit remaining after thinning based on the predicted size and sugar content of the fruit remaining after thinning at the future time point, and determine which of the multiple fruits to thin based on the calculated index value. Effect of the Invention

[0016] According to the present invention, during the fruit tree cultivation process, it is possible for orchard managers and growers to determine which fruits should be thinned based on clear and rational criteria while taking into account the profitability of the fruits. [Brief description of the drawings]

[0017] [Figure 1] 1 is a diagram illustrating an example of a usage environment and configuration of a determination device 100. FIG. [Diagram 2] 2 is a diagram showing an example of an image IM acquired by an image acquisition unit 110. FIG. [Diagram 3] 13 is a diagram showing an example of regression equation data 154A relating to the size of fruit F. FIG. [Figure 4] FIG. 13 is a diagram showing an example of regression equation data 154B regarding the sugar content of fruit F. [Diagram 5]FIG. 13 is a diagram showing another example of regression equation data 154C regarding the sugar content of fruit F. [Figure 6] 13 is a diagram showing an example of a profitability table 156 for fruit F. FIG. [Figure 7] 13 is a diagram showing an example of the expected profit for each fruit thinning pattern calculated by the fruit thinning judgment unit 130. FIG. [Figure 8] 13 is a diagram showing an example of an output result output by a determination result output unit 140. FIG. [Figure 9] 4 is a flowchart showing an example of the flow of processing executed by the determination device 100. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] [overview] Hereinafter, a determination device 100 according to an embodiment of the present invention will be described with reference to the drawings. In the following embodiment, as an example, a determination device 100 for determining which fruits should be thinned in apple orchard cultivation will be described, but the present invention is not limited to apple trees, and can be applied to the cultivation of any crop.

[0019] FIG. 1 is a diagram showing an example of a usage environment and a configuration of the determination device 100. The determination device 100 operates in cooperation with, for example, a camera 10 and a terminal device 20. The camera 10 is, for example, a digital camera using a solid-state imaging element such as a CCD (Charge Coupled Device) or a CMOS (Complementary Metal Oxide Semiconductor), or a smartphone or tablet terminal equipped with a digital camera function. The camera 10 is installed in a position where it can capture an image of an apple fruit F on a tree T in its entirety, and transmits image data of the tree T to the determination device 100 via a network NW. Alternatively, the image data captured by the camera 10 may be manually transmitted to the determination device 100 by a user of the determination device 100 (for example, by connecting the camera 10 and the determination device 100 via a universal serial bus (USB)). In addition, instead of image data capturing an image of the entire fruit tree T, the camera 10 may capture images of multiple different parts of the fruit tree T and transmit the images to the judgment device 100, and the judgment device 100 may obtain an image of the entire fruit tree T by combining the images of the multiple different parts of the fruit tree T that it has received.

[0020] The terminal device 20 is a computer device such as a personal computer, a smartphone, or a tablet terminal. The terminal device 20 communicates with the determination device 100 via a network NW, and as described below, receives the determination results of the determination device 100 regarding fruits that should be thinned, and outputs them on a screen such as a display. This allows a user of the determination device 100 to know which fruits should be thinned and use it as a reference for increasing profits.

[0021] The determination device 100 is a server device such as a web server. The determination device 100 includes, for example, an image acquisition unit 110, a fruit characteristic prediction unit 120, a thinned fruit determination unit 130, a determination result output unit 140, and a storage unit 150. The image acquisition unit 110, the fruit characteristic prediction unit 120, the thinned fruit determination unit 130, and the determination result output unit 140 are each realized by, for example, a hardware processor such as a CPU (Central Processing Unit) executing a program (software). Some or all of these components may be realized by hardware (including circuitry) such as an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a GPU (Graphics Processing Unit), or may be realized by cooperation between software and hardware. The program may be stored in advance in a storage device (a storage device having a non-transient storage medium) such as a hard disk drive (HDD) or a flash memory, or may be stored in a removable storage medium (non-transient storage medium) such as a DVD or a CD-ROM, and installed by mounting the storage medium in a drive device. The storage unit 150 is realized by a storage device such as a HDD, a flash memory, or a random access memory (RAM). The storage unit 150 stores, for example, image data 152, regression equation data 154, and a profitability table 156.

[0022] The image acquisition unit 110 acquires an image IM of a fruit tree T captured by the camera 10 and stores it as image data 152. FIG. 2 is a diagram showing an example of an image IM acquired by the image acquisition unit 110. The image acquisition unit 110 extracts a fruit F included in the image IM by performing a predetermined image processing (for example, template matching or processing using a trained model) on the acquired image IM. In FIG. 2, as an example, fruits F1, F2, and F3 are extracted. At this time, the image acquisition unit 110 defines fruits included in a predetermined range (for example, 1 square meter) among the extracted fruits as one group G, and uses them for subsequent processing. In this embodiment, for convenience of explanation, it is assumed that fruits F1, F2, and F3 are defined as one group included in the predetermined range. In addition, for example, the imaging range of the camera 10 may be set in advance so that the entire image IM coincides with the predetermined range.

[0023] Furthermore, the image acquisition unit 110 estimates the size (horizontal diameter) and the fruit-bearing position (height position from the ground) of the extracted fruit F based on the image IM. The size and the fruit-bearing position of the fruit F captured in the image IM may be estimated using any known technology. In order to estimate the size and the fruit-bearing position of the fruit F with higher accuracy, the fruit tree T may be imaged using multiple cameras 10, and a three-dimensional model of the fruit tree T may be constructed from the images captured by each camera 10 to estimate the size and the fruit-bearing position of the fruit F on the fruit tree T. In FIG. 2, for convenience of explanation, it is assumed that the sizes of the fruits F1, F3, and F2 are estimated in the order of increasing size. Also, it is assumed that the fruit F1 and F3 are in the same fruit-bearing position, while the fruit F2 is estimated to be lower than the fruit F1 and F3.

[0024] The fruit characteristic prediction unit 120 predicts the size and sugar content at a future time point of the fruit F extracted by the image acquisition unit 110 based on the regression equation data 154 stored in the memory unit 150. The regression equation data 154 includes, for example, regression equation data 154A relating to the size of the fruit F, and regression equation data 154B and 154C relating to the sugar content of the fruit F.

[0025] FIG. 3 is a diagram showing an example of regression equation data 154A related to the size of fruit F. In the graph of regression equation data 154A shown in FIG. 3, the horizontal axis represents time (units are, for example, dates) and the vertical axis represents the fruit horizontal diameter (units are, for example, millimeters). Furthermore, in FIG. 3, time point t1 represents the current time in the fruit thinning period, and time point t2 represents the harvest date, which is a future time point. The regression equation data 154A is obtained, for example, by measuring the relationship between time and horizontal diameter of fruits cultivated in the past over a time series and fitting a regression equation.

[0026] The fruit characteristic prediction unit 120 predicts the horizontal diameter of the fruit F at the harvest time t2 by applying the horizontal diameter of the fruit F at the current time t1 estimated by the image acquisition unit 110 to the regression equation data 154A. For example, in the case of fruits F1, F2, and F3 shown in FIG. 2, the fruit characteristic prediction unit 120 obtains a predicted value of the horizontal diameter D(F1, t1) at the harvest time t2 of 80 mm by applying the estimated horizontal diameter D(F1, t1) of the fruit F1 at the current time t1 to the corresponding regression equation RL1. Similarly, the fruit characteristic prediction unit 120 obtains a predicted value of the horizontal diameter D(F2, t2) at the harvest time t2 of 60 mm by applying the estimated horizontal diameter D(F2, t1) of the fruit F2 at the current time t1 to the corresponding regression equation RL3. Similarly, the fruit characteristic prediction unit 120 obtains a predicted value of the horizontal diameter D(F3, t2) = 72 mm at the harvest time t2 by fitting the estimated horizontal diameter D(F3, t1) of fruit F3 at the current time t1 to the corresponding regression equation RL2.

[0027] In addition, when the fruit to be estimated is an apple, a fruit with a relatively large horizontal diameter during the thinning period tends to have a greater degree of growth until the time of harvest, while a fruit with a relatively small horizontal diameter during the thinning period tends to have a smaller degree of growth until the time of harvest. That is, for the multiple regression equations RL shown in Fig. 3, the higher the regression equation RL is located during the thinning period, the steeper the slope from the thinning period to the time of harvest. Therefore, by including multiple regression equations RL corresponding to the horizontal diameter at the current time t1 in the regression equation data 154A, the fruit characteristic prediction unit 120 can estimate the horizontal diameter of the fruit more accurately.

[0028] In another embodiment, the fruit characteristic prediction unit 120 may again obtain the horizontal diameter from an image of the same fruit F at time t1' a predetermined period of time (e.g., one week) after the current time t1, and calculate the slope of a growth curve of the horizontal diameter of the fruit F based on the horizontal diameter at time t1 and the horizontal diameter at time t1'. The fruit characteristic prediction unit 120 estimates the horizontal diameter at harvest time t2 from the horizontal diameter at the current time t1 and the calculated slope of the growth curve. In this way, by using horizontal diameters at multiple times including the current time t1, the horizontal diameter at harvest time t2 can be more accurately estimated.

[0029] Fig. 4 is a diagram showing an example of regression equation data 154B related to the sugar content of fruit F. In the regression equation data 154B shown in Fig. 4, the horizontal axis represents the fruit setting position (units are, for example, centimeters) and the vertical axis represents the sugar content (units are, for example, percentages). The regression equation data 154B is obtained, for example, by measuring the relationship between the fruit setting position and the sugar content at harvest for each fruit cultivated in the past and fitting a regression equation.

[0030] The fruit characteristic prediction unit 120 predicts the sugar content S of the fruit F at the harvest time t2 by fitting the fruit setting position H(F) of the fruit F estimated by the image acquisition unit 110 to the regression equation data 154B. For example, in the case of the fruits F1, F2, and F3 shown in FIG. 2, the fruit characteristic prediction unit 120 obtains the sugar content S(F1)=14% at the harvest time t2 by fitting the estimated fruit setting position H(F1) of the fruit F1 to the regression equation. Similarly, the fruit characteristic prediction unit 120 obtains the sugar content S(F2)=13% at the harvest time t2 by fitting the estimated fruit setting position H(F2) of the fruit F2 to the regression equation. Similarly, the fruit characteristic prediction unit 120 obtains the sugar content S(F3)=14% at the harvest time t2 by fitting the estimated fruit setting position H(F3) of the fruit F3 to the regression equation. In this way, the fruit characteristic prediction unit 120 predicts that the higher the fruit F is located in the fruit F, the higher the sugar content S will be at the harvest time t2.

[0031] In this embodiment, as an example, the sugar content of each fruit at the harvest time t2 is predicted using a regression equation that represents the relationship between fruit setting position and sugar content, but any parameter that generally has a correlation with sugar content may be used instead of the fruit setting position. For example, the daily sunshine hours for each fruit estimated from the image IM may be used as a parameter instead of the fruit setting position. More specifically, the fruit characteristic prediction unit 120 may predict using a regression equation that represents the relationship between sunshine hours and sugar content, assuming that the fruit F with longer sunshine hours has a higher sugar content S.

[0032] Fig. 5 is a diagram showing another example of regression equation data 154C relating to the sugar content of fruit F. In the graph of regression equation data 154C shown in Fig. 5, the horizontal axis represents fruit density (units, for example, number per square meter) and the vertical axis represents sugar content (units, for example, percentage). The regression equation data 154C is obtained, for example, by measuring the relationship between the average sugar content and the number of fruits of multiple fruits grouped every square meter among fruits cultivated in the past, and fitting a regression equation.

[0033] The fruit characteristic prediction unit 120 predicts the average sugar content of group G of fruit F at harvest time t2 by applying the regression equation data 154C to group G of fruit F grouped by the image acquisition unit 110. For example, in the case of FIG. 5, since group G including fruits F1, F2, and F3 has a fruit density of 3, the fruit characteristic prediction unit 120 predicts that the average sugar content of group G at harvest time t2 is 14%. On the other hand, if any fruit is thinned from group G including fruits F1, F2, and F3, the fruit density of group G becomes 2, and in this case, the average sugar content of group G at harvest time t2 is predicted to be 15%. In this way, the fruit characteristic prediction unit 120 predicts that the lower the fruit density of fruit F, the higher the sugar content S at harvest time t2.

[0034] The fruit characteristic prediction unit 120 finally predicts the sugar content of each fruit F as the average value of the sugar content of each fruit F predicted based on the fruiting position and the average sugar content of the group G to which each fruit F belongs. For example, in the case of fruits F1, F2, and F3 shown in Fig. 2, the fruit characteristic prediction unit 120 predicted the sugar contents of fruits F1, F2, and F3 to be 14%, 13%, and 14%, respectively, based on the fruiting position, while predicting the average sugar content of group G to be 14%, and therefore predicts the final sugar contents of fruits F1, F2, and F3 to be (14+14) / 2=14%, (13+14) / 2=13.5%, and (14+14) / 2=14%, respectively.

[0035] At this time, the fruit characteristic prediction unit 120 predicts the sugar content in each pattern in which one or more fruits are thinned from a plurality of fruits. For example, the fruit characteristic prediction unit 120 predicts the final sugar content of fruits F1, F2, and F3 for each pattern in which fruits F1, F2, and F3 are thinned from group G (a total of eight patterns including a pattern in which no fruits are thinned). For example, when fruit F2 is thinned from group G, as described above, the fruit density of group G is 2, so the average sugar content of group G is 15%, and the fruit characteristic prediction unit 120 predicts the sugar content of fruits F1 and F3 to be (14+15) / 2=14.5% and (14+15) / 2=14.5%, respectively.

[0036] In this embodiment, for ease of explanation, as an example, the fruit characteristic prediction unit 120 predicts the sugar content of each fruit as the average value of the sugar content predicted based on the fruit position and the sugar content predicted based on the fruit density. However, the present invention is not limited to such a configuration, and it is sufficient to predict the sugar content of each fruit by taking into account at least one of the sugar content predicted based on the fruit position and the sugar content predicted based on the fruit density.

[0037] The thinning fruit determination unit 130 refers to the profitability table 156 for fruit F and calculates the expected profit of the fruit left in each thinning pattern based on the horizontal diameter and sugar content of fruit F predicted by the fruit characteristic prediction unit 120. The thinning fruit determination unit 130 determines the fruit F to be thinned according to the pattern that has the maximum expected profit out of the expected profits calculated for each pattern.

[0038] 6 is a diagram showing an example of the profitability table 156 for fruit F. The profitability table 156 includes, for example, a profitability table 156A for diameter and a profitability table 156B for sugar content. The thinning fruit determination unit 130 calculates the expected profit for each fruit F by comparing the diameter and sugar content of fruit F predicted by the fruit characteristic prediction unit 120 with the profitability table 156A and the profitability table 156B, respectively.

[0039] For example, in the case of fruits F1, F2, and F3 shown in Fig. 2, the thinning fruit judgment unit 130 uses the expected horizontal diameter of fruit F1 (80 mm) and the expected sugar content of fruit F1 (14%) as keys to compare with the profitability table 156A and the profitability table 156B, and calculates the expected profit of fruit F1 as 50 yen x 2 = 100 yen. Similarly, the thinning fruit judgment unit 130 uses the expected horizontal diameter of fruit F2 (60 mm) and the expected sugar content of fruit F2 (13.5%) as keys to compare with the profitability table 156A and the profitability table 156B, and calculates the expected profit of fruit F2 as 30 x 1.5 = 45 yen. Similarly, the thinning fruit judgment unit 130 uses the expected horizontal diameter of fruit F3 (72 mm) and the expected sugar content of fruit F3 (14%) as keys to compare with the profitability table 156A and the profitability table 156B, and calculates the expected profit of fruit F3 as 40 x 2 = 80 yen. That is, the thinned fruit determination unit 130 calculates the expected profit for group G in the pattern in which fruits F1, F2, and F3 are not thinned as 100+45+80=225 yen.

[0040] On the other hand, if fruit F2 is thinned from group G, the predicted sugar content of fruits F1 and F3 will each be 14.5%, as described above. Therefore, the thinning fruit determination unit 130 calculates the predicted profit for group G in the pattern in which fruit F2 is thinned as 50 yen x 3 + 40 x 3 = 270 yen. This means that the predicted profit for group G in the pattern in which fruit F2 is thinned is higher than the predicted profit for group G in the pattern in which fruits F1, F2, and F3 are not thinned.

[0041] In this way, the thinning fruit determination unit 130 calculates the expected profit for all patterns (8 patterns) in which at least one of fruits F1, F2, and F3 is thinned, and identifies the pattern with the maximum expected profit among all 8 patterns. Figure 7 is a diagram showing an example of the expected profit for each thinning pattern calculated by the thinning fruit determination unit 130. As shown in Figure 7, the expected profit of 270 yen for group G in the pattern in which fruit F2 is thinned is the maximum. Therefore, the thinning fruit determination unit 130 determines fruit F2 as fruit F that should be thinned.

[0042] 6, regarding the size of fruit F, the profitability table 156A relating to the horizontal diameter is used to identify the sales per fruit F. However, the present invention is not limited to such a configuration, and for example, the profitability table 156A linking the weight (unit: grams, for example) of fruit F with the unit price may be used to identify the sales per fruit F. In that case, the storage unit 150 may also store a mapping table linking the horizontal diameter and weight of the fruit.

[0043] Furthermore, in the above explanation, for the sake of convenience, a method for determining which fruits to thin out is described when three fruits F1, F2, and F3 are contained within a specified range (one square meter); however, the above-mentioned algorithm can be applied even if the specified range and the number of fruits contained therein are any range and number.

[0044] The judgment result output unit 140 outputs the judgment result obtained by the thinning fruit judgment unit 130 to the terminal device 20. FIG. 8 is a diagram showing an example of the output result output by the judgment result output unit 140. As shown in FIG. 8, the judgment result output unit 140 transmits the image IM captured by the camera 10 and information MS representing the fruit judged by the thinning fruit judgment unit 130 to be thinned to the terminal device 20 via the network NW, and the terminal device 20 outputs the received information on a screen such as a display. This allows the user of the judgment device 100 to know which fruits should be thinned and use it as a reference for increasing profits. In addition to the information shown in FIG. 8, the judgment result output unit 140 may display a specific profit amount (225 yen → 270 yen) that will be increased by thinning.

[0045] Next, a process flow executed by the determination device 100 will be described with reference to Fig. 9. Fig. 9 is a flowchart showing an example of a process flow executed by the determination device 100. The process of the flowchart shown in Fig. 9 is executed when an image of a plurality of apples captured by the camera 10 during a fruit thinning period is received, for example.

[0046] First, the image acquisition unit 110 of the determination device 100 acquires an image of multiple fruits and extracts each fruit contained in the image by performing a predetermined image processing on the acquired image (step S100). Next, the fruit feature prediction unit 120 of the determination device 100 predicts the size and sugar content of the fruit remaining after thinning at a future time point for each pattern in which one or more fruits are thinned from the multiple fruits (step S102).

[0047] Next, the thinned fruit judgment unit 130 of the judgment device 100 calculates the expected profit of the fruit remaining after thinning based on the size and sugar content predicted by the fruit characteristic prediction unit 120 (step S104). Next, the thinned fruit judgment unit 130 identifies a fruit thinning pattern that maximizes the expected profit, and judges the fruit to be thinned according to the identified fruit thinning pattern (step S106). Next, the judgment result output unit 140 outputs the judgment result by the thinned fruit judgment unit 130 to the terminal device 20 (step S108). This ends the processing related to this flowchart.

[0048] According to the present embodiment described above, images of multiple fruits of a crop are acquired, the size and sugar content of the fruit remaining after thinning at a future time point are predicted based on the acquired images for each pattern in which one or more fruits are thinned from the multiple fruits, an index value relating to the profitability of the fruit remaining after thinning is calculated based on the predicted size and sugar content of the fruit remaining after thinning at a future time point, and a determination is made of which of the multiple fruits should be thinned based on the calculated index value. This allows orchard managers and growers to determine which fruits should be thinned based on clear and reasonable criteria during the fruit cultivation process, while taking into account the profitability of the fruit.

[0049] The above describes the form for carrying out the present invention using an embodiment, but the present invention is not limited to such an embodiment, and various modifications and substitutions can be made within the scope that does not deviate from the gist of the present invention. [Explanation of symbols]

[0050] 10. Camera 20 Terminal Equipment 100 Judgment device 110 Image acquisition unit 120 Fruit Characteristics Prediction Unit 130 Fruit Thinning Judgment Department 140 Judgment result output unit 150 Storage section

Claims

1. an acquisition unit that acquires images of a plurality of fruits of a crop; a prediction unit that predicts the size and sugar content of a fruit remaining after thinning at a future time point for each pattern in which one or more fruits are thinned from the plurality of fruits based on the image; a determination unit that calculates an index value relating to profits of the fruits remaining after the thinning based on the predicted size and sugar content of the fruits remaining after the thinning at the future time point, and determines which of the plurality of fruits to thin based on the calculated index value. Judgment device.

2. The acquisition unit acquires a plurality of images of the plurality of fruits in time series, the prediction unit predicts a size of the fruit remaining after the fruit thinning at the future time point based on a thickening rate of each of the plurality of fruits captured in the plurality of images. The determination device according to claim 1 .

3. The prediction unit predicts the sugar content of the fruit remaining after the fruit thinning at the future time point based on the fruit setting positions of each of the plurality of fruits. The determination device according to claim 1 .

4. the prediction unit predicts a higher sugar content of the fruit remaining after the fruit thinning at the future time point as the fruit-setting position of each of the plurality of fruits is higher; The determination device according to claim 3 .

5. The prediction unit predicts the sugar content of the fruit remaining after the fruit thinning at the future time point based on the fruit setting density of the plurality of fruits. The determination device according to claim 1 .

6. The prediction unit predicts a higher sugar content of the fruit remaining after the fruit thinning at the future time point as the fruit-bearing density of the plurality of fruits is lower. The determination device according to claim 5 .

7. the determining unit calculates an expected profit of the fruit remaining after the fruit thinning as the index value, and determines which of the plurality of fruits to thin so as to maximize the calculated expected profit. The determination device according to claim 1 .

8. The crop is apple. The determination device according to claim 1 .

9. The computer Acquire images of multiple fruits of a crop; predicting, based on the images, the size and sugar content of the fruit remaining after the thinning at a future time point for each pattern in which one or more fruits are thinned from the plurality of fruits; calculating an index value relating to profit for the fruits remaining after the fruit thinning based on the predicted size and sugar content of the fruits remaining after the fruit thinning at the future time point, and determining which of the plurality of fruits to thin based on the calculated index value; Judgment method.

10. On the computer, acquiring images of a plurality of fruits of a crop; predicting, based on the images, the size and sugar content of a fruit remaining after thinning at a future time point for each pattern in which one or more fruits are thinned from the plurality of fruits; calculating an index value relating to profits of the fruits remaining after the fruit thinning based on the predicted size and sugar content of the fruits remaining after the fruit thinning at the future time point, and determining which of the plurality of fruits to thin based on the calculated index value; program.

Images