Vehicle

Abstract

Provided is a vehicle capable of efficiently spraying a pesticide. The vehicle is provided with a photographing device that photographs a control target around the vehicle, a control device that determines whether or not the pesticide needs to be sprayed, and a spraying device that sprays the pesticide, so that the estimation and control of the disease and pests can be simultaneously performed. In addition, the photographing device photographs the lowermost end to the uppermost end in the vertical direction among the control targets around the vehicle, so that the estimation and control of the disease and pests can be simultaneously performed as the vehicle even in the case of spraying the pesticide to a tall control target such as a fruit tree. Therefore, the pesticide can be efficiently sprayed.

Description

TECHNICAL FIELD

[0001] The present application relates to a vehicle that sprays a pesticide to an object to be controlled. BACKGROUND

[0002] A system that sprays a pesticide to an object to be controlled is known. For example, a support system for agriculture described in Patent Literature 1 is such a system. In this Patent Literature 1, a support system for agriculture that is provided with a photographing device, a control device, and a spraying device is disclosed. The photographing device of Patent Literature 1 is a device that is provided to an unmanned flight body such as a multicopter and photographs inside a farm. The control device of Patent Literature 1 is a device that calculates a growth state of a crop inside a farm using an image obtained by photographing with the photographing device, estimates occurrence of a disease or pest based on a result of the calculation, and creates a work plan related to spraying of a pesticide based on a result of the estimation. The spraying device of Patent Literature 1 is a device that sprays a pesticide to a farm in which a disease or pest is estimated to have occurred.

[0003] PRIOR ART DOCUMENTS

[0004] PATENT LITERATURE

[0005] Patent Literature 1: Japanese Patent Application Publication No. 2021-106554 SUMMARY

[0006] In a system that sprays a pesticide to an object to be controlled, it is necessary to have a function of photographing the object to be controlled, a function of judging whether or not it is necessary to spray a pesticide, and a function of spraying a pesticide, and the like. In the support system for agriculture described in Patent Literature 1, each device that realizes each of the above functions is provided separately. In such a support system for agriculture, after photographing the entire area of a farm that becomes an object, it is judged whether or not it is necessary to spray a pesticide for each of the areas inside the farm, and a pesticide is sprayed to an area in which it is necessary according to a result of the judgment. There is room for improvement in spraying a pesticide. In addition, in a case where the object to be controlled is a tall object such as a fruit tree, when a photographing position in a vertical direction is set uniformly, there is a possibility that a pesticide cannot be sprayed efficiently.

[0007] The present application has been achieved in light of the above circumstances, and aims to provide a vehicle that can spray a pesticide efficiently.

[0008] The gist of the first application is to provide (a) a vehicle that sprays a pesticide to an object to be controlled, in which (b) a photographing device photographs the object to be controlled around the vehicle, (c) a control device judges whether or not it is necessary to spray the pesticide based on an image when the object to be controlled is photographed, and (d) a spraying device sprays the pesticide to the object to be controlled in which it is judged that it is necessary to spray the pesticide, (e) the photographing device photographs a lowermost end to an uppermost end in a vertical direction among the object to be controlled around the vehicle.

[0009] Further, the other application is the vehicle described in the first application, wherein the control device determines whether or not the pesticide needs to be sprayed by applying data of the image for determining whether or not the pesticide needs to be sprayed to a learning model determined in advance, wherein the learning model is realized by supervised learning based on machine learning, and represents a relationship between data of the image and whether or not the pesticide needs to be sprayed.

[0010] According to the first application, the vehicle for spraying the pesticide to the object to be controlled is provided with the imaging device for imaging the object to be controlled around the vehicle, the control device for determining whether or not the pesticide needs to be sprayed, and the spraying device for spraying the pesticide, so that the estimation and the control of the disease and the pest can be simultaneously performed. Further, the lowermost end to the uppermost end in the vertical direction among the object to be controlled around the vehicle is imaged by the imaging device, so that the estimation and the control of the disease and the pest can be simultaneously performed as the vehicle even in the case where the pesticide is sprayed to the tall object to be controlled such as the fruit tree. Therefore, the pesticide can be efficiently sprayed.

[0011] Further, according to the other application, whether or not the pesticide needs to be sprayed is determined by applying data of the image for determining whether or not the pesticide needs to be sprayed to a learning model determined in advance, wherein the learning model is realized by supervised learning based on machine learning, and represents a relationship between data of the image and whether or not the pesticide needs to be sprayed. Thus, whether or not the pesticide needs to be sprayed can be determined in accordance with a visual difference close to human determination. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1 is a diagram illustrating a schematic structure of the vehicle to which the present application is applied, and is a diagram illustrating an appearance of the vehicle in work.

[0013] Figure 2 is a diagram illustrating a schematic structure of the vehicle to which the present application is applied, and is a diagram illustrating main parts of various control functions and control systems in the vehicle.

[0014] Figure 3 is a diagram showing one example of a color tone difference in a predetermined range of an object image.

[0015] Figure 4 (a) of is a diagram showing one example of an object image of a predetermined range in which it is determined that the pesticide needs to be sprayed, and (b) is a diagram showing one example of a corrected object image.

[0016] Figure 5 is a diagram showing one example of a learning model.

[0017] Figure 6This is a flowchart illustrating a main portion of the control operation of the electronic control device, and is a flowchart illustrating the control operation for efficiently spreading pesticides.

[0018] (Explanation of Symbols)

[0019] 10: Vehicle; 12: Camera; 14: Electronic control unit (control unit); 16: Spreading device; 20: Pesticide; 30: Learning model; 102: Object to be sprayed; PIC: Image; PICb: Predetermined reference image; PICs: Object image (image used to determine whether pesticide spraying is necessary); PICsc: Corrected object image (corrected image). DETAILED DESCRIPTION

[0020] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0021] [Example]

[0022] Figure 1 as well as Figure 2 Each of them is a diagram for explaining a schematic structure of a vehicle 10 to which the present invention is applied.

[0023] Figure 1 It is a diagram illustrating the appearance of the vehicle 10 during operation. Figure 2 1 and 2 are diagrams for explaining various control functions and main parts of a control system in the vehicle 10 .

[0024] The vehicle 10 is, for example, a well-known electric vehicle capable of automatic driving. The vehicle 10 is driven automatically to navigate a predetermined path within the farm 100 (see Figure 1 The vehicle 10 is a vehicle that sprays pesticides 20 onto pest control objects 102 in a farm 100. The farm 100 is a farm such as an orchard. The pest control objects 102 are crops such as fruit trees planted in the farm 100. Such crops include, for example, leaves in addition to fruits. The pest control objects 102 are planted, for example, in rows at predetermined intervals. The pest control objects 102 are crops that are the targets of prevention and extermination of pests and diseases. Pests and diseases are diseases and pests of crops. The pesticide 20 is a well-known agricultural agent. In addition, the vehicle 10 may be an engine vehicle, a vehicle that is driven by a person through automatic driving, or a vehicle that can be driven manually.

[0025] The vehicle 10 includes an imaging device 12 , an electronic control device 14 , a spreading device 16 , and the like.

[0026] The imaging device 12 is provided, for example, at the front in the advancing and retreating direction of the vehicle 10. The imaging device 12 is, for example, a monocular camera that images the surroundings of the vehicle 10. The imaging device 12 images the control target object 102 of the surroundings of the vehicle 10, particularly, the front and the side. The imaging device 12 outputs information of the object image PICs as image information Ipic to the electronic control device 14. The object image PICs are images PICs of the control target object 102 imaged at the time of dispensing the pesticide 20, and are images PICs used for determining whether or not the pesticide 20 needs to be dispensed (refer to (a) of the description to be given later). Figure 4

[0027] The dispensing device 16 is provided, for example, at the rear in the advancing and retreating direction of the vehicle 10. The dispensing device 16 is provided, for example, with a tank 22 that stores the pesticide 20, and a sprayer 24 that sprays the pesticide 20, and the like. The dispensing device 16 dispenses the pesticide 20 against the control target object 102.

[0028] The electronic control device 14 is a control device, that is, a controller of the vehicle 10, associated with the control of the travel of the vehicle 10, the imaging device 12, the dispensing device 16, and the like. The electronic control device 14 is configured, for example, to include a so-called microcomputer provided with a CPU, a RAM, a ROM, an input / output interface, and the like. The CPU performs various controls of the vehicle 10 by performing signal processing in accordance with a program stored in advance in the ROM while using the temporary storage function of the RAM.

[0029] The image information Ipic is supplied from the imaging device 12 to the electronic control device 14. In addition, various signals based on detection values detected by various sensors and the like (for example, a vehicle speed sensor 26 and the like) provided in the vehicle 10 (for example, a vehicle speed V and the like) are supplied to the electronic control device 14, respectively. A photographing instruction signal Spic for imaging the control target object 102 of the surroundings of the vehicle 10 is output from the electronic control device 14 to the imaging device 12. In addition, a dispensing instruction signal Sspr for dispensing the pesticide 20 against the control target object 102 is output from the electronic control device 14 to the dispensing device 16. In addition, a control instruction signal for automatic driving is output from the electronic control device 14 to various devices not shown provided in the vehicle 10.

[0030] ​The imaging device 12 images the control target object 102 around the vehicle 10 according to an imaging instruction signal Spic from the electronic control device 14, and outputs image information Ipic including information of the object image PICs to the electronic control device 14. The electronic control device 14 determines whether or not the pesticide 20 needs to be sprayed, based on the object image PICs included in the acquired image information Ipic. The electronic control device 14 outputs a spraying instruction signal Sspr for spraying the pesticide 20 against the control target object 102 corresponding to the object image PICs included in the acquired image information Ipic to the spraying device 16, in a case where it is determined that the pesticide 20 needs to be sprayed. The spraying device 16 sprays the pesticide 20 against the control target object 102 determined to need the pesticide 20, according to the spraying instruction signal Sspr from the electronic control device 14.

[0031] However, in a case where the control target object 102 is a tall object such as a fruit tree, when the vertical direction position imaged by the imaging device 12 is set to a uniform height, there is a possibility that the pesticide 20 cannot be efficiently sprayed.

[0032] Therefore, the imaging device 12 is provided with a detection sensor 28 such as a laser radar or an infrared sensor, or the like, which detects the control target object 102 around the vehicle 10. The imaging device 12 detects the vertical direction length or height of the control target object 102 around the vehicle 10 by the detection sensor 28 while the vehicle 10 is running at the time of spraying the pesticide 20. For example, the imaging device 12 detects the position where the control target object 102 exists by the detection sensor 28 while the vehicle 10 is running at the time of spraying the pesticide 20. Then, the imaging device 12 images the control target object 102 from the lowermost end to the uppermost end in the vertical direction while the vehicle 10 is running at the time of spraying the pesticide 20.

[0033] The higher the vehicle speed V while the vehicle 10 is running, the more the object image PICs are likely to become unclear, and it is difficult to determine whether or not the pesticide 20 needs to be sprayed. The vehicle 10 preferably runs at a predetermined vehicle speed range Vpic at which the object image PICs imaged by the imaging device 12 are clear, at the time of spraying the pesticide 20. In particular, in order to improve the efficiency at the time of spraying the pesticide 20, it is preferable to run at the highest vehicle speed Vpicmax in the vehicle speed range Vpic.

[0034] On the other hand, the higher the vehicle speed V while the vehicle 10 is running, the more difficult it is to spray the pesticide 20 against the control target object 102. The vehicle 10 preferably runs at a predetermined vehicle speed range Vspr at which the pesticide 20 is appropriately sprayed by the spraying device 16, at the time of spraying the pesticide 20. In particular, in order to improve the efficiency at the time of spraying the pesticide 20, it is preferable to run at the highest vehicle speed Vsprmax in the vehicle speed range Vspr.

[0035] In the vehicle speed range in which it is possible to determine whether or not the pesticide 20 needs to be spread, and in which it is possible to appropriately spread the pesticide 20, in order to spread the pesticide 20 with the greatest efficiency, the lower one of the maximum vehicle speed Vpicmax and the maximum vehicle speed Vsprmax can be selected. That is, the vehicle 10 travels at the vehicle speed Vmin, which is the lower one of the maximum vehicle speed Vpicmax and the maximum vehicle speed Vsprmax, as an upper limit when the pesticide 20 is spread (refer to Fig. 2). For example, the electronic control device 14 outputs a control command signal for automatically driving with the vehicle speed Vmin as an upper limit to each of the devices provided in the vehicle 10, which are not shown. Figure 1 ). For example, the electronic control device 14 outputs a control command signal for automatically driving with the vehicle speed Vmin as an upper limit to each of the devices provided in the vehicle 10, which are not shown.

[0036] The determination of whether or not the pesticide 20 needs to be spread by the electronic control device 14 will be described in detail. For example, the electronic control device 14 performs predetermined image processing on the object image PICs, and calculates a color tone difference Dt of the color tone of the control object 102 (particularly, the leaves) in the predetermined range AR and the color tone at normal times. As the color tone at normal times, for example, a predetermined color tone, an average color tone of the control object 102 in the farm 100, and an average color tone of the control object 102 outside the predetermined range AR of the object image PICs are used. The electronic control device 14 determines whether or not the pesticide 20 needs to be spread in accordance with whether or not the color tone difference Dt in the predetermined range AR of the object image PICs is equal to or greater than a necessity determination value THn. The meaning of determining whether or not the pesticide 20 needs to be spread in accordance with whether or not the color tone difference Dt is equal to or greater than the necessity determination value THn is the same as that of inferring whether or not the disease and the pest have occurred in accordance with whether or not the color tone difference Dt is equal to or greater than the necessity determination value THn. The necessity determination value THn is, for example, a predetermined necessity determination value for determining that the color tone difference Dt is large to the extent that the pesticide 20 needs to be spread. The electronic control device 14 determines the predetermined range AR in which the color tone difference Dt is equal to or greater than the necessity determination value THn as a range in which the pesticide 20 needs to be spread.

[0037] Figure 3 is a view showing one example of the color tone difference Dt in the predetermined range AR of the object image PICs. Figure 4 (a) of is a view showing one example of the object image PICs in which the predetermined range AR in which the pesticide 20 is determined to be necessary is determined. In Figure 3 In the third predetermined range AR3, the color tone difference Dt becomes equal to or greater than the necessity determination value THn in (b) of Figure 4 As shown in (a) of, the third predetermined range AR3 is determined as a range in which the pesticide 20 needs to be spread.

[0038] The situation ST at the time when the object 102 is imaged by the imaging device 12 is not the same. The situation ST at the time of imaging has, for example, a plurality of kinds of brightness of the surroundings of the vehicle 10, a kind of the object 102, a growth condition of the object 102, a health state of the object 102, a place of the farm 100, and the like. The brightness of the surroundings of the vehicle 10 is, for example, brightness due to a difference in season, time zone, weather condition, shadow, and the like.

[0039] The electronic control device 14 stores a predetermined reference image PICb that is an image PIC at the time when the object 102 is imaged by the imaging device 12 under a predetermined situation STf. The electronic control device 14 corrects the object image PICs in accordance with a difference between the predetermined situation STf at the time when the object 102 that becomes the basis of the predetermined reference image PICb is imaged and the situation STs at the time when the object 102 that becomes the basis of the object image PICs is imaged. For example, the electronic control device 14 corrects the object image PICs in accordance with a difference between the above-described situation STs and the above-described predetermined situation STf by a predetermined image correction process. In the present embodiment, the corrected object image PICs is referred to as a corrected object image PICsc (see (b) of FIG. 6). The electronic control device 14 determines whether or not the pesticide 20 needs to be spread in accordance with the corrected object image PICsc. For example, the electronic control device 14 determines whether or not the pesticide 20 needs to be spread in accordance with whether or not a difference in hue Dt in a predetermined range AR of the corrected object image PICsc is equal to or greater than a necessity determination value THn. In this case, the electronic control device 14 can store only the predetermined situation STf required for correction of the object image PICs in advance, and does not need to store the predetermined reference image PICb. Alternatively, the electronic control device 14 can determine whether or not the pesticide 20 needs to be spread in accordance with a difference between the predetermined reference image PICb and the corrected object image PICsc. For example, the electronic control device 14 can determine whether or not the pesticide 20 needs to be spread in accordance with whether or not a difference in hue Dt between a hue in the predetermined reference image PICb and a hue in the corrected object image PICsc is equal to or greater than the necessity determination value THn. Figure 4

[0040] When the electronic control device 14 determines whether or not the pesticide 20 needs to be spread, a learning model 30 based on machine learning can be used, for example. For example, the electronic control device 14 determines whether or not the pesticide 20 needs to be spread by applying the object image PICs to the learning model 30. The learning model 30 is a predetermined learned model that indicates a relationship between data of the image PIC and whether or not the pesticide 20 needs to be spread. The learning model 30 is realized by supervised learning based on machine learning in which data of the image PIC and whether or not the pesticide 20 needs to be spread are used as training data. Whether or not the pesticide 20 needs to be spread can be replaced with presence or absence of occurrence of a disease or pest. ​

[0041] Figure 5 is a diagram showing an example of the learning model 30. Figure 5 In the example, the learning model 30 is a neural network based on the data of the image PIC and the state ST when the pest control object 102 that forms the basis of the image PIC is photographed. The learning model 30 is a model that can be constructed by modeling the neural cell group of an organism through software that can be implemented using a computer program, or through hardware composed of a combination of electronic components. The learning model 30 is a multi-layer structure including an input layer composed of i neural cell elements (= neurons) Pi1, an intermediate layer composed of j neural cell elements Pj2, and an output layer composed of k neural cell elements Pk3. The intermediate layer can also be a multi-layer structure. In addition, in the learning model 30, in order to transmit the state of the neural cell elements from the input layer to the output layer, a transmission element Dij with a weighted value Wij and a transmission element Djk with a weighted value Wjk are provided. The transmission element Dij is a transmission element that connects the i neural cell elements Pi1 and the j neural cell elements Pj2, respectively. The transmission element Djk is a transmission element that connects the j neuron elements Pj2 and the k neuron elements Pk3.

[0042] The learning model 30 is a necessity determination system that determines whether or not the pesticide 20 needs to be sprayed. This necessity determination system is also a pest estimation system. In the learning model 30, weighted values ​​Wij and Wjk are machine-learned using a predetermined algorithm. In supervised learning in the learning model 30, training data, i.e., training signals, determined in the vehicle 10 are used. Figure 5 In the learning model 30, as training signals for the input layer, for example, data of an image PIC and data of a state ST when the pest control object 102 is photographed are provided. Figure 5 In the learning model 30, as the training signal for the output layer, the determination result of whether the pesticide 20 needs to be sprayed, that is, the inference result of whether there are pests and diseases, is provided. Figure 5 The learning model 30 uses the target image PICs and the state STs when the pest control object 102 serving as the basis of the target image PICs is photographed to determine whether the pesticide 20 needs to be applied. Figure 5 In the case of the learning model 30, the situation ST when the pest control target 102 is photographed is also learned, so there is no need to correct the target image PICs.

[0043] The learning model 30 can also be a predetermined learned model that represents a relationship of the data of the image PIC, whether or not the pesticide 20 needs to be applied, the kind of the disease and pest that occurs in the object 102, and the degree of development of the damage due to the disease and pest. That is, the learning model 30 can also represent a relationship of the data of the image PIC and the kind of the disease and pest that occurs in the object 102 and the degree of development of the damage due to the disease and pest. In this case, as the determination result for the output layer composed of the k neural cell elements Pk3 in Figure 5 the training signal, in addition to the determination result of whether or not the pesticide 20 needs to be applied, the determination result of the kind of the disease and pest that occurs in the object 102 and the degree of development of the damage due to the disease and pest is provided. The kind of the disease and pest is, for example, a disease name such as "downy mildew" and the like. The degree of development of the damage due to the disease and pest is, for example, "small", "medium", "large". For example, in the output layer in which it is determined that the pesticide 20 needs to be applied, the determination result of the kind of the disease and pest and the degree of development of the damage due to the disease and pest is provided. The electronic control device 14 judges whether or not the pesticide 20 needs to be applied by applying the object images PICs to the learning model 30, and determines the kind of the disease and pest and the degree of development of the damage due to the disease and pest. The electronic control device 14 decides the kind of the pesticide 20 and the required amount in accordance with the determined kind of the disease and pest and the degree of development of the damage due to the disease and pest. The kind of the pesticide 20 is, for example, Bordeaux mixture corresponding to "downy mildew". The required amount of the pesticide 20 is, for example, the dilution ratio and the application amount. In the application instruction signal Sspr output from the electronic control device 14, in addition to the pesticide 20 to be applied to the object 102, the kind of the pesticide 20 and the required amount are included. The application device 16 applies the pesticide 20 corresponding to the disease and pest to the object 102 determined to require the pesticide 20 in the required amount from the electronic control device 14 in accordance with the application instruction signal Sspr.

[0044] The electronic control device 14 can also use the learning model 30 to estimate the tendency of the occurrence timing, occurrence site, and the like of the disease and pest (refer to the portion enclosed with a single-dot chain line B in Figure 1 ). The electronic control device 14 stores the estimation result of the tendency as data. The electronic control device 14 can also use the estimation result to correct, for example, the application amount of the pesticide 20 at the time of the next application of the pesticide 20. Thus, it is possible to grasp the tendency of the disease and pest that occurs in the object 102. In addition, it is possible to add the concentration and dilution to the application amount of the pesticide 20.

[0045] Figure 6 is a flowchart that explains the main part of the control operation of the electronic control device 14, and is a flowchart that explains the control operation for efficiently applying the pesticide 20, and is repeatedly executed, for example, during traveling.

[0046] In Figure 6In the present embodiment, in step (hereinafter, the step is omitted) S10, the object image PICs included in the image information Ipic from the imaging device 12 is acquired. Next, in S20, it is judged whether or not the pesticide 20 needs to be sprayed based on the object image PICs. In the case where the judgment of this S20 is negative, the present routine ends. In the case where the judgment of this S20 is affirmative, in S30, the spraying instruction signal Sspr for spraying the pesticide 20 against the object of control 102 corresponding to the acquired object image PICs is output to the spraying device 16.

[0047] As described above, according to the present embodiment, the imaging device 12, the electronic control device 14, and the spraying device 16 are provided on the vehicle 10, so that the estimation and the control of the pest can be performed simultaneously. The vehicle 10 is not a flying body, so that, for example, the energy consumption accompanying the increase in the mounting weight of the vehicle 10 is suppressed. In addition, the lowermost end to the uppermost end in the vertical direction in the object of control 102 is imaged by the imaging device 12, so that even in the case where the pesticide 20 is sprayed against the tall object of control 102 such as a fruit tree, the estimation and the control of the pest can be performed simultaneously as the vehicle 10. At this time, it is possible to find, for example, the pest occurring at a lower position in the tall object of control 102. Therefore, the pesticide 20 can be sprayed efficiently.

[0048] In addition, according to the present embodiment, the vehicle 10 is caused to travel with the vehicle speed Vmin, which is the lower one of the maximum vehicle speed Vpicmax and the maximum vehicle speed Vsprmax, as the upper limit at the time of spraying the pesticide 20. Thereby, the imaging of the clear object image PICs in which it is possible to judge whether or not the pesticide 20 needs to be sprayed and the spraying of the appropriate pesticide 20 can be performed with the maximum efficiency. The vehicle 10 is not a flying body, so that, for example, the energy efficiency is also good in the work in the lower speed range.

[0049] In addition, according to the present embodiment, the object image PICs is corrected by the electronic control device 14 based on the difference between the predetermined situation STf and the situation STs at the time of imaging the object of control 102 which becomes the basis of the object image PICs. In addition, whether or not the pesticide 20 needs to be sprayed is judged by the electronic control device 14 based on the corrected object image PICsc. Thereby, it is possible to acquire the object image PICs which is not easily affected by the situation STs such as the brightness of the surroundings of the vehicle 10 and the kind of the object of control 102, and the estimation and the control of the pest can be performed appropriately.

[0050] In addition, according to the present embodiment, whether or not the pesticide 20 needs to be sprayed is judged by applying the object image PICs to the learning model 30 by the electronic control device 14, so that it is possible to judge whether or not the pesticide 20 needs to be sprayed based on the visual difference close to the human judgment method.

[0051] In addition, according to the present embodiment, by applying the object images PICs to the learning model 30 using the electronic control device 14, the type of the pest and the degree of damage due to the pest are determined, and the type and the amount of the pesticide 20 are decided in accordance with the type of the pest and the degree of damage due to the pest. Thus, the type and the dilution ratio of the pesticide effective against the pest are selected. In addition, in the case where the pest and the type of the pesticide 20 mounted on the vehicle 10 do not match, for example, the spreading of the pesticide 20 can be suspended to reduce the amount of use of the pesticide 20.

[0052] The above describes the embodiment of the present application in detail with reference to the drawings, but the present application can be applied in other ways.

[0053] For example, in the above-described embodiment, it is determined whether or not the pesticide 20 needs to be spread using the color difference Dt, but the present application is not limited to this. For example, the growth state or the health state of the object 102 to be controlled, or the like can be used to determine whether or not the pesticide 20 needs to be spread. In this way, various methods can be used when determining whether or not the pesticide 20 needs to be spread.

[0054] In addition, in the above-described embodiment, the photographing device 12 is provided with the detection sensor 28, and the length or the height in the vertical direction of the object 102 to be controlled is detected, but the present application is not limited to this. For example, the photographing device 12 can process the image PIC to recognize the photographing range, and photograph the lowermost end to the uppermost end in the vertical direction of the object 102 to be controlled. That is, the photographing device 12 does not necessarily need to detect the length or the height in the vertical direction of the object 102 to be controlled in the periphery of the vehicle 10. In this case, the photographing device 12 does not necessarily need to be provided with the detection sensor 28. In any case, the photographing device 12 can at least photograph the lowermost end to the uppermost end in the vertical direction of the object 102 to be controlled when the vehicle 10 is running at the time of spreading the pesticide 20.

[0055] In addition, in the above-described embodiment, the spreading device 16 is provided at the rear of the vehicle 10, but the present application is not limited to this. For example, the spreading device 16 can be provided at a carriage connected to the rear of the vehicle 10 body in the advancing and retreating direction. The carriage is a vehicle that is towed by the vehicle 10 body and runs integrally with the vehicle 10 body, and constitutes a part of the vehicle 10. Therefore, the spreading device 16 provided at the carriage corresponds to a device provided at the vehicle 10.

[0056] In addition, in the above-described embodiment, the photographing device 12 is not limited to being provided at the front of the vehicle 10, and the spreading device 16 is not limited to being provided at the rear of the vehicle 10. In any case, it is only necessary that the structure be such that the pesticide 20 can be spread with respect to the object 102 to be controlled determined to need the pesticide 20.

[0057] In addition, in the above-described embodiment, in the case where the learning model 30 is a predetermined learned model that represents the relationship between the data of the image PIC and whether or not the agricultural chemical 20 needs to be sprayed, as Figure 5 the training signal for the input layer in the above-described embodiment does not need to include data of the situation ST.

[0058] Furthermore, the above-described is only one embodiment, and the present application can be implemented in a manner that various changes and improvements are applied by those skilled in the art.

Claims

1. A vehicle (10) for spreading pesticide (20) onto an object (102) to be controlled, characterized in that: The vehicle (10) comprises: A photographing device (12) for photographing the objects to be removed (102) around the vehicle (10); A control device (14) determines whether the pesticide (20) needs to be spread based on an image of the pest control object (102); and The spreading device (16) spreads the pesticide (20) on the pest control object (102) determined to require the pesticide (20) to be spread. The photographing device (12) photographs the vertical direction of the object to be removed (102) around the vehicle (10), from the lowest end to the highest end. When spreading the pesticide (20), the vehicle is driven with the lower of the maximum vehicle speed in the predetermined vehicle speed range that makes the image used to determine whether the pesticide (20) needs to be spread clear and the maximum vehicle speed in the predetermined vehicle speed range that allows the pesticide (20) to be properly spread by the spreading device (16) as the upper limit.

2. The vehicle (10) according to claim 1, characterized in that The control device (14) corrects the image based on the difference between the predetermined condition when the pest control object (102) serving as the basis for the predetermined reference image was photographed and the condition when the pest control object (102) serving as the basis for the image for judging whether the pesticide (20) needs to be sprayed, and judges whether the pesticide (20) needs to be sprayed based on the corrected image.

3. The vehicle (10) according to claim 1 or 2, characterized in that The control device (14) determines whether the pesticide (20) needs to be spread by applying the image data used to determine whether the pesticide (20) needs to be spread to a predetermined learning model (30), wherein the learning model (30) is implemented by supervised learning based on machine learning and represents the relationship between the image data and whether the pesticide (20) needs to be spread.

4. The vehicle (10) according to claim 3, characterized in that The learning model (30) further represents the relationship between the image data, the type of pests and diseases occurring in the control object (102), and the degree of damage caused by the pests and diseases. The control device (14) determines the type of the pest and disease and the degree of development of the disaster by applying the image data used to determine whether the pesticide (20) needs to be spread to the learning model (30), and determines the type and required amount of the pesticide (20) based on the type of the pest and disease and the degree of development of the disaster.

Images