Travel route management system, travel route management program, recording medium recording travel route management program, travel route management method

Abstract

A travel path management system (A) for a work vehicle (1) that travels for work in a ridge field having a plurality of ridge portions formed of piled-up soil and a trench-shaped portion provided between two ridge portions adjacent to each other, the travel path management system (A) includes an acquisition unit (40) that acquires ridge information related to at least one of the ridge portions and the trench-shaped portion, the acquisition unit (40) is configured to acquire the ridge information of a portion in the ridge field located ahead of a travel direction of the work vehicle (1), and the travel path management system (A) includes an inference unit (23) that infers a ridge direction, which is an extension direction of the ridge portions, based on the ridge information acquired by the acquisition unit (40), and a path generation unit (24) that generates a target travel path of the work vehicle (1) based on the ridge direction inferred by the inference unit (23).

Description

Technical Field

[0001] This invention relates to a travel path management system for a work vehicle used for working in ridged fields. Background Technology

[0002] As for work vehicles that operate in fields, the content described in Patent Document 1 is already well known. This work vehicle (referred to as a "combine harvester" in Patent Document 1) is configured to automatically move in the field based on signals received from GPS satellites, and also possesses a grain quantity detection means for detecting the amount of grain in the grain bin. Furthermore, when the detection value detected by the grain quantity detection means reaches or exceeds a set value, in order to discharge grain from the grain bin, the work vehicle interrupts the harvesting operation and automatically moves towards the vicinity of the truck.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent Application Publication No. 2001-69836 Summary of the Invention

[0006] The problem that the invention will solve

[0007] Patent document 1 does not describe a structure suitable for driving in ridge-laying operations.

[0008] The purpose of this invention is to provide a driving path management system capable of generating suitable target driving paths in ridges.

[0009] Methods for solving problems

[0010] The present invention is characterized by a travel path management system for a work vehicle operating in a ridged field, the ridged field having a plurality of ridges formed by piled soil and a ditch between two adjacent ridges. The travel path management system includes a ridge information acquisition unit that acquires information related to at least one of the ridges and the ditch, the acquisition unit being configured to acquire the ridge information of the portion of the ridge located ahead of the travel direction of the work vehicle in the ridged field. The system includes an inference unit and a path generation unit. The inference unit infers the extension direction of the ridges, i.e., the ridge direction, based on the ridge information acquired by the acquisition unit. The path generation unit generates a target travel path for the work vehicle based on the ridge direction inferred by the inference unit.

[0011] In this invention, the inference unit infers the ridge direction based on ridge information of the portion of the ridge located ahead of the direction of travel of the work vehicle in the ridged area. Then, a target driving path is generated based on the inferred ridge direction. Thus, a driving path management system capable of generating suitable target driving paths in the ridged area can be realized.

[0012] Furthermore, in this invention, it is preferable that the acquisition unit acquires the ridge information over time, the inference unit updates the inference result of the ridge direction over time based on the ridge information acquired by the acquisition unit, and the path generation unit updates the target driving path over time based on the inference result updated by the inference unit.

[0013] When viewed from above, the ridges extend straight, and when the work vehicle travels along the ridge direction, the direction of extension of the portion of the ridge in front of the work vehicle's direction of travel remains constant throughout the vehicle's movement. However, in reality, when viewed from above, the ridges sometimes curve and sometimes serpentine. Therefore, the direction of extension of the portion of the ridge in front of the work vehicle's direction of travel is not necessarily constant during the work vehicle's movement.

[0014] Here, based on the above structure, the inferred ridge direction is updated over time, and simultaneously, the target travel path is updated over time based on the updated inference. Therefore, even when the ridges are sometimes curved and sometimes serpentine when viewed from above, it is possible to achieve a structure that updates the inferred ridge direction based on changes in the ridge direction ahead of the work vehicle's travel direction and appropriately changes the target travel path.

[0015] Furthermore, in this invention, it is preferred that the acquisition unit is a photographing device that takes pictures of the area in the ridge located in front of the direction of travel of the work vehicle, the ridge information is the photographed image acquired by the photographing device, and the inference unit infers the ridge direction based on the color information contained in the photographed image.

[0016] According to this structure, compared to structures that infer ridge direction without relying on color information contained in the captured image, the inference unit can more accurately infer the ridge direction. Therefore, the generated target driving path is more likely to be a suitable target driving path.

[0017] Furthermore, in this invention, it is preferred that the inference unit divides the analysis object region within the captured image into a first region corresponding to the ridge and a second region corresponding to the groove based on the color information.

[0018] According to this structure, the extension direction of the first region indicates the ridge direction. Furthermore, since the extension direction of the furrow is essentially the same as the extension direction of the ridge, the extension direction of the second region also indicates the ridge direction. Therefore, according to this structure, by calculating either the extension direction of the first region or the extension direction of the second region, the inference unit can easily and accurately infer the ridge direction. Consequently, the generated target driving path can be easily adapted.

[0019] Furthermore, in this invention, it is preferred that the inference unit infers the ridge direction based on the ridge information obtained by the acquisition unit by simultaneously calculating the average value of each of the various directions in which the plurality of ridges extend, or the various directions in which the plurality of furrows extend, or the various directions in which one or more ridges and one or more furrows extend.

[0020] In a structure where the inference unit calculates only the extension direction of a single ridge and infers the ridge direction by determining that direction as the ridge direction, it is conceivable that the inference accuracy of the ridge direction would deteriorate in a ridged field if only that ridge is skewed.

[0021] Furthermore, in a structure where the inference unit only calculates the extension direction of a single groove and infers the ridge direction by defining that direction as the ridge direction, it is also conceivable that the accuracy of the ridge direction inference may decrease.

[0022] Here, based on the above structure, the inference unit calculates each direction of the extension of multiple ridges, or each direction of the extension of multiple furrows, or each direction of the extension of one or more ridges and one or more furrows. Furthermore, the inference unit infers the ridge direction by calculating the average value of the calculated directions.

[0023] Therefore, the situation where the accuracy of ridge direction inference deteriorates, as described above, can be avoided. Thus, based on the above structure, a travel path management system that minimizes the deterioration of ridge direction inference accuracy can be implemented through the inference unit.

[0024] Furthermore, in this invention, it is preferred that the inference unit infers the ridge direction by calculating the direction of extension of the second region in the captured image, the path generation unit generates the target driving path, such that the work vehicle operates along the extension direction of the groove corresponding to the second region, and the work vehicle operates with its driving device grounded relative to the groove corresponding to the second region.

[0025] According to this structure, if the work vehicle travels along the target travel path, it travels along the ridge direction, and the travel device of the work vehicle is grounded relative to the ditch. Therefore, the work vehicle can travel in a stable posture.

[0026] Furthermore, in this invention, preferably, when there are multiple second regions in the analysis target area, the inference unit determines the second region with the largest area among the multiple second regions as the target region, and calculates the direction of extension of the target region in the captured image, thereby inferring the ridge direction. The path generation unit generates the target driving path, so that the work vehicle operates along the extension direction of the groove-shaped portion corresponding to the target region, and the work vehicle operates with the driving device grounded relative to the groove-shaped portion corresponding to the target region.

[0027] According to this structure, if the work vehicle travels along the target travel path, the vehicle's travel device is grounded relative to the trench-shaped portion corresponding to the second region with the largest area among the multiple second regions existing in the analysis area. Therefore, compared to the case where the travel device is grounded relative to the trench-shaped portion corresponding to the second region with a narrower area, it is less likely that the grounding surface of the travel device will extend from the trench-shaped portion and contact the ridge. Thus, the posture of the work vehicle is more easily stabilized.

[0028] Furthermore, in this invention, preferably, the acquisition unit acquires the ridge information over time, the inference unit updates the inference result of the ridge direction over time based on the ridge information acquired by the acquisition unit, the path generation unit updates the target driving path over time based on the inference result updated by the inference unit, and after determining the target area, the inference unit narrows the analysis object area to increase the proportion of the target area in the analysis object area.

[0029] After the inference unit determines the target area, in a structure where the analysis object area remains unchanged, it is conceivable that the size relationship of multiple second areas existing within the analysis object area changes as the work vehicle moves. In this case, the second area identified as the target area changes as the work vehicle moves.

[0030] For example, if there are two second regions, left and right, within the analysis target area, and at a certain moment the second region on the right has a larger area than the second region on the left, then at that moment, the second region on the right is identified as the target region. Subsequently, if the second region on the left has a larger area than the second region on the right, then in this state, the second region on the left is identified as the target region. That is, in this case, the second region identified as the target region changes from the second region on the right to the second region on the left.

[0031] Therefore, the groove-shaped portion that is grounded relative to the traveling device changes as the work vehicle moves. Furthermore, when the groove-shaped portion that is grounded relative to the traveling device changes, the traveling device will run over the ridge. Thus, it is conceivable that the work vehicle will vibrate.

[0032] Here, based on the above structure, after determining the target area, the inference unit narrows the analysis object area to increase the proportion of the target area within the analysis object area. Therefore, the narrowed analysis object area is less likely to include a second area other than the second area identified as the target area. Thus, a travel path management system can be implemented that makes it difficult for the second area identified as the target area to change as the work vehicle moves.

[0033] Another feature of the present invention is a path management program for a work vehicle operating in a ridged field, the ridged field having a plurality of ridges composed of piled soil and ditch-like sections between two adjacent ridges. The program implements a ridge information acquisition function via computer, which acquires information related to at least one of the ridges and the ditch-like sections. The acquisition function is configured to acquire the ridge information of the portion of the ridge located in front of the work vehicle in the direction of travel of the work vehicle in the ridged field. The program also implements an inference function and a path generation function via computer. The inference function infers the extension direction of the ridge based on the ridge information acquired by the acquisition function, i.e., the ridge direction. The path generation function generates a target travel path for the work vehicle based on the ridge direction inferred by the inference function.

[0034] Another feature of the present invention is a recording medium for recording a path management program for a work vehicle used in working on a ridged field, the ridged field having a plurality of ridges composed of piled soil and furrows between two adjacent ridges. A computer is used to implement a ridge information acquisition function, which acquires information related to at least one of the ridges and the furrows. This acquisition function is configured to acquire the ridge information of the portion of the ridge located ahead of the work vehicle's direction of travel in the ridged field, and to record the path management program. The path management program enables a computer-based inference function and a path generation function. The inference function infers the direction of extension of the ridges, i.e., the ridge direction, based on the ridge information acquired by the acquisition function, and the path generation function generates a target travel path for the work vehicle based on the ridge direction inferred by the inference function.

[0035] Another feature of the present invention is a method for managing the travel path of a work vehicle operating in a ridged field, wherein the ridged field has a plurality of ridges composed of piled soil and a ditch between two adjacent ridges. The method includes a step of obtaining ridge information, i.e., obtaining information related to at least one of the ridges and the ditch. In the obtaining step, the ridge information of the portion of the ridge located in front of the travel direction of the work vehicle in the ridged field is obtained. The method includes an inference step and a path generation step. The inference step infers the direction of extension of the ridge, i.e., the ridge direction, based on the ridge information obtained by the obtaining step. The path generation step generates a target travel path of the work vehicle based on the ridge direction inferred by the inference step. Attached Figure Description

[0036] Figure 1 This is a left-side view of a combine harvester.

[0037] Figure 2 It is a top view showing the ridged field and the combine harvester.

[0038] Figure 3 It is a top view showing the positional relationship between the groove section and the left and right tracks.

[0039] Figure 4 This is a rear view showing the positional relationship between the groove and the left and right tracks.

[0040] Figure 5 It is a block diagram representing the structure related to the control unit.

[0041] Figure 6 This is a diagram showing an example of a photographed image obtained by a photographing device.

[0042] Figure 7 It is a diagram representing the area of ​​the object to be analyzed within the captured image.

[0043] Figure 8 This is a diagram representing the first and second regions.

[0044] Figure 9 It is a graph representing the target area and an approximate straight line.

[0045] Figure 10 This is a rear view showing the positional relationship between the groove and the left and right tracks.

[0046] Figure 11 This is a diagram illustrating an example where the target driving path changes due to an update of the target driving path.

[0047] Figure 12 This is a diagram illustrating the inference of the ridge direction in the first other embodiment.

[0048] Figure 13 This is a diagram showing the case where the analysis target area is reduced in other implementations (1). Detailed Implementation

[0049] Embodiments of the present invention will be described with reference to the accompanying drawings. Furthermore, in the following description, unless otherwise stated, Figure 1 as well as Figure 3 The direction of arrow F is set to "forward", and the direction of arrow B is set to "backward". Figure 3 , Figure 4 , Figure 10 The direction of arrow L is set to "left", and the direction of arrow R is set to "right". Additionally, [the following is a separate, unrelated section:] Figure 1 , Figure 4 , Figure 10 The direction of arrow U is set to "up", and the direction of arrow D is set to "down".

[0050] [The overall structure of a combine harvester]

[0051] like Figure 1 As shown, the conventional combine harvester 1 (equivalent to the "operating vehicle" of the present invention) includes a harvesting section H, left and right tracks 11, a driving section 12, a threshing device 13, a grain box 14, a conveying section 16, a grain discharge device 18, and a satellite positioning module 80.

[0052] Left and right tracks 11 are located on the lower part of the combine harvester 1. Furthermore, the left and right tracks 11 are driven by power from an engine (not shown) mounted on the combine harvester 1. The combine harvester 1 is capable of self-propelled movement via the left and right tracks 11.

[0053] Additionally, the driver's cab 12, threshing device 13, and grain bin 14 are located on the upper side of the left and right tracks 11. An operator can sit in the driver's cab 12 to monitor the operation of the combine harvester 1. Alternatively, the operator can also monitor the operation of the combine harvester 1 from outside the machine.

[0054] The driving unit 12 includes a driver's seat 12a and a cockpit 12b. The driver's seat 12a is located inside the cockpit 12b. The operator can sit in the driver's seat 12a.

[0055] The grain discharge device 18 is located on the upper side of the grain bin 14. In addition, the satellite positioning module 80 is mounted on the upper surface of the driver's unit 12.

[0056] The harvesting section H is located at the front of the combine harvester 1. Furthermore, the conveying section 16 is located at the rear of the harvesting section H. The harvesting section H includes a cutting device 15 and a wheel 17.

[0057] The harvesting device 15 harvests rice stalks from the field. Additionally, the wheel 17, driven by rotating around its axle 17b along the left-right direction of the machine body, feeds the rice stalks to be harvested. The harvested rice stalks from the harvesting device 15 are then conveyed to the conveyor section 16.

[0058] According to this structure, the harvesting section H harvests the grain in the field. Furthermore, the combine harvester 1 is capable of cutting movement, that is, it moves along the left and right tracks 11 while cutting the stalks of the field using the cutting device 15.

[0059] The harvested rice stalks, cut by the harvesting section H, are conveyed to the rear of the machine via the conveyor section 16. From there, the harvested rice stalks are transported to the threshing unit 13.

[0060] In the threshing device 13, the harvested rice stalks are threshed. The resulting rice grains are stored in a grain bin 14. The rice grains stored in the grain bin 14 are discharged from the machine as needed via a grain discharge device 18.

[0061] like Figure 2 As shown, the combine harvester 1 performs harvesting travel in the ridge field FI (equivalent to the "operational travel" of this invention).

[0062] In addition, such as Figure 1 , Figure 3 , Figure 4 As shown, the combine harvester 1 has a left track 11, namely the left track 11L, and a right track 11, namely the right track 11R (equivalent to the "travel device" of the present invention).

[0063] [Regarding ridge land]

[0064] like Figures 2 to 4 As shown, in this embodiment, the ridged land FI has multiple ridges 31 and multiple trenches 32. The ridges 31 are composed of piled-up soil. The trenches 32 are trench-shaped portions provided between two adjacent ridges 31.

[0065] In other words, the ridge FI has multiple ridges 31 made of piled soil and ditch-like sections 32 located between two adjacent ridges 31.

[0066] like Figure 4 As shown, soybeans 33 are planted on the ridge 31 in this embodiment. However, the present invention is not limited to this, and other types of crops besides soybeans 33 may also be planted on the ridge 31.

[0067] In addition, such as Figure 2 as well as Figure 3 As shown, the harvesting width of the harvesting section H in this embodiment corresponds to the sum of the widths of the three ridges 31 and the two grooves 32 located between these ridges 31.

[0068] Here, the combine harvester 1 is configured to be able to follow the path of... Figure 5 The target driving path LI generated by the path generation unit 24 is shown in the figure (refer to the path generation unit 24). Figure 3 The system will drive automatically. Furthermore, the target driving path LI will be determined by the driving path management system A (refer to...). Figure 5 The system manages the combine harvester 1, which travels along the ridges FI.

[0069] The following is a detailed description of the driving route management system A.

[0070] [Structure of the Driving Route Management System]

[0071] like Figure 4 As shown, the combine harvester 1 includes a control unit 20. Furthermore, the control unit 20 is included in the travel path management system A. The control unit 20 includes a vehicle position calculation unit 21 and a travel control unit 22.

[0072] like Figure 1 As shown, the satellite positioning module 80 receives GPS signals from the artificial satellite GS, which is used for GPS (Global Positioning System). Furthermore, as... Figure 5 As shown, the satellite positioning module 80 sends the positioning data representing the position of the combine harvester 1 to the vehicle position calculation unit 21 based on the received GPS signal.

[0073] Furthermore, the present invention is not limited thereto. The satellite positioning module 80 may also not utilize GPS. For example, the satellite positioning module 80 may also utilize GNSS other than GPS (GLONASS, Galileo, michibiki, BeiDou, etc.).

[0074] The vehicle position calculation unit 21 calculates the position coordinates of the combine harvester 1 based on the positioning data output by the satellite positioning module 80. The calculated position coordinates of the combine harvester 1 are then sent to the driving control unit 22.

[0075] In addition, such as Figure 5 As shown, the combine harvester 1 includes a camera 40 (equivalent to the "acquisition unit" of the present invention). Furthermore, the camera 40 is included in the travel path management system A.

[0076] Furthermore, the control unit 20 includes an inference unit 23 and a path generation unit 24. Both the inference unit 23 and the path generation unit 24 are included in the driving path management system A.

[0077] In this embodiment, the imaging device 40 is a camera (e.g., a CCD camera or a CMOS camera). Figure 1As shown, the camera device 40 is mounted on the upper part of the left front of the cockpit 12b. Thus, the camera device 40 is positioned at the center of the front of the combine harvester 1 in the left-right direction.

[0078] The camera device 40 is positioned facing the front of the combine harvester 1. Thus, as... Figure 2 As shown, the imaging device 40 captures images of the area FA located in front of the combine harvester 1 in the direction of travel within the ridge FI. Thus, the imaging device 40 obtains images of the ridge 31 and the furrow 32.

[0079] According to this structure, the imaging device 40 is able to acquire ridge information. Furthermore, ridge information refers to information related to at least one of the ridge portion 31 and the furrow portion 32. In this embodiment, the ridge information is the captured image obtained by the imaging device 40.

[0080] Thus, the travel path management system A includes a camera device 40, which acquires ridge information related to at least one of the ridge section 31 and the furrow section 32. Furthermore, the camera device 40 is configured to acquire ridge information of the portion of the ridged field FI located ahead of the combine harvester 1 in the direction of travel.

[0081] Alternatively, the imaging device 40 may be configured to capture images of only the ridge 31 and the furrow 32, or it may be configured to capture images of only the furrow 32.

[0082] like Figure 5 As shown, the captured image obtained by the imaging device 40 is sent to the inference unit 23.

[0083] The inference unit 23 infers the ridge direction based on the captured image received from the imaging device 40. Furthermore, the ridge direction refers to the extension direction of the ridge 31. However, the extension direction of the furrow 32 is essentially the same as the extension direction of the ridge 31, so the extension direction of the furrow 32 can also be treated as the "ridge direction".

[0084] That is, the driving path management system A has an inference unit 23, which infers the extension direction of the ridge 31, i.e. the ridge direction, based on the captured image obtained by the imaging device 40.

[0085] like Figure 5 As shown, the inference result of the inference unit 23 is sent to the path generation unit 24.

[0086] The path generation unit 24 generates the target travel path LI of the combine harvester 1 based on the inference result of the inference unit 23 (refer to...). Figure 3 That is, the driving path management system A has a path generation unit 24, which generates a target driving path L1 for the combine harvester 1 based on the ridge direction inferred by the inference unit 23.

[0087] like Figure 5 As shown, information about the target driving path LI generated by the path generation unit 24 is sent to the driving control unit 22.

[0088] The driving control unit 22 is configured to control the left and right tracks 11. Furthermore, the driving control unit 22 controls the automatic driving of the combine harvester 1 based on the position coordinates of the combine harvester 1 received from the vehicle position calculation unit 21 and information representing the target driving path L1 received from the path generation unit 24. More specifically, as... Figure 3 As shown, the driving control unit 22 controls the left and right tracks 11, enabling them to perform cutting motion by automatically traveling along the target driving path LI.

[0089] In this embodiment, the travel control unit 22 controls the left and right tracks 11, causing the combine harvester 1 to travel in a state where the satellite positioning module 80 is located on the target travel path LI when viewed from above. However, the present invention is not limited to this. The travel control unit 22 may also be configured to control the left and right tracks 11, causing the combine harvester 1 to travel in a state where a predetermined part of the combine harvester 1 other than the satellite positioning module 80 is located on the target travel path LI when viewed from above.

[0090] In addition, the control unit 20 and the vehicle position calculation unit 21 included in the control unit 20 can be physical devices such as microcomputers, or functional units in software.

[0091] [Inference regarding the direction of the ridge]

[0092] The following is a detailed explanation of the ridge direction inference performed by the inference unit 23.

[0093] Figure 6 The combine harvester 1 is shown in Figure 3 This is an example of capturing images taken by the imaging device 40 during the driving process in the shown state. Figure 6 As shown, the image captured by the imaging device 40 shows the area FA in the ridge FI located in front of the combine harvester 1 in the direction of travel (refer to...). Figure 2 The groove 32 and soybean 33 are shown in the image. Although not shown, the ridge 31 is located below the soybean 33, so the ridge 31 can also be captured in the image obtained by the imaging device 40.

[0094] In addition, such as Figure 6 As shown, the image captured by the imaging device 40 also captured the cutting device 15 and the wheel 17.

[0095] like Figure 6As shown, the inference unit 23 determines the analysis object region 50 within the captured image. Furthermore, the method used to determine the analysis object region 50 is not particularly limited; the position and size of the analysis object region 50 can be determined using a neural network based on machine learning, or the analysis object region 50 can be determined according to a pre-set position and size.

[0096] In this embodiment, the analysis target area 50 is rectangular. However, the present invention is not limited to this, and the analysis target area 50 may have any shape.

[0097] Figure 7 yes Figure 6 A magnified view of the analysis object region 50 shown. (See attached image.) Figure 6 as well as Figure 7 As shown, the area FA located in front of the combine harvester 1 in the direction of travel within the ridge FI of the analysis object area 50 was photographed (refer to...). Figure 2 The groove 32 and soybean 33 are shown in the diagram. Although not shown here, the ridge 31 is located below the soybean 33, so the ridge 31 can also be captured in the analysis area 50.

[0098] like Figure 8 As shown, the inference unit 23 divides the analysis object region 50 into a first region 51 and a second region 52 based on the color information contained in the captured image.

[0099] Detailed explanation, such as Figure 7 as well as Figure 8 As shown, within the analysis object area 50, the portion of soybean 33 photographed is divided into a first area 51 based on color information. Furthermore, since the ridge 31 is located below the soybean 33, the first area 51 corresponds to the ridge 31.

[0100] Additionally, the portion of the analysis object region 50 outside the first region 51 is designated as the second region 52. The second region 52 corresponds to the groove-like portion 32.

[0101] That is, based on color information, the inference unit 23 divides the analysis object region 50 in the captured image into a first region 51 corresponding to the ridge 31 and a second region 52 corresponding to the groove 32.

[0102] Here, the inference unit 23 is configured to determine the target region 53, which is the second region 52 with the largest area among the multiple second regions 52, when there are multiple second regions 52 in the analysis object region 50.

[0103] exist Figure 8 In the example shown, there are four second regions 52 within the analysis object region 50. Therefore, the inference unit 23 compares the areas of each second region 52, such as... Figure 9As shown, target region 53 is determined. In this example, among the four second regions 52, the rightmost second region 52 has the largest area. Therefore, the rightmost second region 52 is determined as target region 53.

[0104] In addition, Figure 9 In the example shown, the second region 52, other than the second region 52 identified as target region 53, is deleted. Additionally, the first region 51 is also deleted. However, the invention is not limited to this; the second region 52 and the first region 51, other than the second region 52 identified as target region 53, may also be retained.

[0105] like Figure 9 As shown, the inference unit 23 calculates the center point 54 of the target region 53 in the left-right direction of the analysis object region 50. At this time, the inference unit 23 calculates the center points 54 of multiple positions in the up-down direction of the analysis object region 50.

[0106] Then, the inference unit 23 uses, for example, the least squares method to approximate the calculated multiple center points 54 with an approximate straight line 55. Thus, the inference unit 23 calculates the approximate straight line 55.

[0107] The approximate line 55 is a line that represents the position of the target region 53 in the analysis object region 50 and the direction of extension of the target region 53 in the analysis object region 50. That is, the calculation of the approximate line 55 is equivalent to the calculation of the direction of extension of the target region 53 in the captured image.

[0108] like Figure 5 As shown, the inference unit 23 is configured to obtain the position coordinates of the combine harvester 1 from the vehicle position calculation unit 21. In addition, based on the detection results of the inertial measurement device (not shown) equipped on the combine harvester 1 and the position coordinates of the combine harvester 1, the inference unit 23 can calculate the attitude orientation of the combine harvester 1.

[0109] Furthermore, based on the position coordinates and orientation of the combine harvester 1, the inference unit 23 converts the position and direction of the approximate straight line 55 in the analysis area 50 into the top-view position and direction of the ridge FI. Thus, as... Figure 3 as well as Figure 9 As shown, the inference unit 23 converts the approximate straight line 55 into an inference line 56. In other words, the inference unit 23 calculates the inference line 56 based on the position coordinates and orientation of the combine harvester 1 and the approximate straight line 55.

[0110] like Figure 3 As shown, the extension direction of the inference line 56 is consistent with the extension direction of the groove 32. That is, the extension direction of the inference line 56 corresponds to the ridge direction. Furthermore, the inference line 56 is located at the center of the width direction of the groove 32. Additionally, in... Figure 3 In the middle, the groove-shaped portion 32 where the inference line 56 is located corresponds to Figure 9 The target area shown is 53.

[0111] Here, the calculation of the approximate straight line 55 and the calculation of the inferred line 56 are equivalent to inferring the extension direction of the furrow 32 corresponding to the target region 53. Therefore, the calculation of the approximate straight line 55 and the calculation of the inferred line 56 are equivalent to inferring the ridge direction. In addition, the extension direction of the approximate straight line 55 and the extension direction of the inferred line 56 are equivalent to the ridge direction inferred by the inference unit 23. Furthermore, the extension direction of the approximate straight line 55 and the extension direction of the inferred line 56 are equivalent to the "inference result" of the present invention.

[0112] That is, the inference unit 23 infers the ridge direction based on the color information contained in the captured image. Additionally, the inference unit 23 infers the ridge direction by calculating the direction in which the second region 52 extends in the captured image. Furthermore, when there are multiple second regions 52 in the analysis target region 50, the inference unit 23 determines the second region 52 with the largest area among the multiple second regions 52 as the target region 53, and simultaneously infers the ridge direction by calculating the direction in which the target region 53 extends in the captured image.

[0113] like Figure 5 As shown, the inference unit 23 sends information representing the inference line 56 to the path generation unit 24. The path generation unit 24 generates a target driving path LI based on the information representing the inference line 56. More specifically, as... Figure 3 As shown, the path generation unit 24 generates a target driving path LI such that the inference line 56 and the target driving path LI are parallel to each other, and the target driving path LI is located at a predetermined distance D1 from the inference line 56 to the left side of the machine.

[0114] In addition, the predetermined distance D1 is equivalent to the distance between the center position of the right track 11R in the left-right direction of the aircraft and the center position of the satellite positioning module 80.

[0115] According to this structure, the target travel path LI extends along the ridge direction. More specifically, the target travel path LI extends along the extension direction of the groove 32 corresponding to the target region 53.

[0116] like Figure 5 As shown, information about the target travel path LI generated by the path generation unit 24 is sent to the travel control unit 22. Furthermore, as described above, the travel control unit 22 controls the left and right tracks 11, enabling them to perform cut-and-go travel by automatically traveling along the target travel path LI. At this time, the travel control unit 22 controls the left and right tracks 11 to travel in a state where the satellite positioning module 80 is positioned on the target travel path LI when viewed from above.

[0117] According to this structure, such as Figure 3 As shown, the combine harvester 1 travels along the extension direction of the trench 32 corresponding to the target area 53. Additionally, as... Figure 3 as well as Figure 4 As shown, the combine harvester 1 moves with its right track 11R grounded to the groove 32 corresponding to the target area 53.

[0118] Thus, the path generation unit 24 generates a target travel path LI, causing the combine harvester 1 to travel along the extension direction of the trench 32 corresponding to the second region 52, and the combine harvester 1 travels with its right track 11R grounded relative to the trench 32 corresponding to the second region 52. Additionally, the path generation unit 24 generates a target travel path LI, causing the combine harvester 1 to travel along the extension direction of the trench 32 corresponding to the target region 53, and the combine harvester 1 travels with its right track 11R grounded relative to the trench 32 corresponding to the target region 53.

[0119] Therefore, as Figure 4 as well as Figure 10 As shown, the machine body's posture is stable because the right track 11R is grounded relative to the groove 32. As a result, the position of the cutting device 15 with the cutter 15a is stably maintained in the proper position relative to the soybean 33.

[0120] In addition, Figure 4 In the example shown, not only the right track 11R, but also the left track 11L is grounded relative to the groove 32. However, in Figure 10 In the example shown, the left track 11L has opened onto the ridge 31.

[0121] The combine harvester 1 in this embodiment includes a lifting device (not shown) that allows the main body of the machine to roll by changing the height position of the main body relative to the left and right tracks 11. Such lifting devices are well known, so a detailed description of the lifting device mechanism is omitted.

[0122] like Figure 10 In the example shown, even when the left track 11L is on the ridge 31, the main body of the machine can be rolled by the lifting device to keep the main body of the machine horizontal, thereby keeping the cutter 15a horizontal. Therefore, even when the left track 11L is on the ridge 31, the position of the cutter 15a can be stably maintained at an appropriate position relative to the soybean 33.

[0123] [Update regarding the target driving route]

[0124] The following describes the updating of the target travel path LI when the combine harvester 1 automatically travels along the target travel path LI.

[0125] In this embodiment, the imaging device 40 acquires images at predetermined time intervals. That is, the imaging device 40 acquires information over time.

[0126] Furthermore, whenever an image is captured, the capturing device 40 sends the captured image to the inference unit 23. And whenever an image is received from the capturing device 40, the inference unit 23 updates the inference result for the ridge direction. That is, the inference unit 23 updates the inference result for the ridge direction over time based on the captured image acquired by the capturing device 40.

[0127] Furthermore, whenever the inference result of the ridge direction is updated, the inference unit 23 in this embodiment sends the updated inference result to the path generation unit 24. And whenever the inference result of the ridge direction is received from the inference unit 23, the path generation unit 24 updates the target travel path LI. That is, the path generation unit 24 updates the target travel path LI over time based on the inference result updated by the inference unit 23.

[0128] Figure 11 The diagram illustrates an example where the target travel path LI changes due to an update of the target travel path LI. In this example, the combine harvester 1 performs a harvesting journey, first passing through a first position P1. Furthermore, at the moment the combine harvester 1 reaches the first position P1, an image is acquired by the imaging device 40. Also, at the moment the combine harvester 1 reaches the first position P1, the target travel path LI has not yet been generated.

[0129] Based on the captured image obtained by the imaging device 40, such as Figure 11 As shown in the lower part of the paper, the inference unit 23 determines the target area 53 and simultaneously calculates an approximate straight line 55. Furthermore, the inference unit 23 calculates an inference line 56. Additionally, the inference line 56 calculated when the combine harvester 1 reaches the first position P1 is used as the first inference line 56a.

[0130] In this case, the path generation unit 24 generates a target travel path LI based on the first inference line 56a. The target travel path LI generated at this time is taken as the first path LI1. After the first path LI1 is generated, the combine harvester 1 automatically performs harvesting travel along the first path LI1.

[0131] Afterwards, combine harvester 1 reaches the second position P2. At the moment when combine harvester 1 reaches the second position P2, an image is captured by the imaging device 40.

[0132] Based on the captured image obtained by the imaging device 40, such as Figure 11As shown in the center of the paper in the vertical direction, the inference unit 23 determines the target area 53 and updates the approximate straight line 55 by recalculating it. Furthermore, based on the updated approximate straight line 55, the inference unit 23 updates the inference line 56 by recalculating it. Thus, the inference result of the ridge direction made by the inference unit 23 is updated. Additionally, the inference line 56 calculated when the combine harvester 1 reaches the second position P2 is used as the second inference line 56b.

[0133] In this case, the path generation unit 24 updates the target travel path LI based on the second inference line 56b by regenerating the target travel path LI. The target travel path LI generated at this time is taken as the second path LI2. At this time, the target travel path LI changes from the first path LI1 to the second path LI2. After the second path LI2 is generated, the combine harvester 1 automatically performs harvesting travel along the second path LI2.

[0134] Afterwards, combine harvester 1 reaches the third position P3. At the moment when combine harvester 1 reaches the third position P3, an image is captured by the imaging device 40.

[0135] Based on the image captured by the shooting device 40, such as Figure 11 As shown on the upper part of the paper, the inference unit 23 determines the target region 53 and updates the approximate straight line 55 by recalculating it. Furthermore, based on the updated approximate straight line 55, the inference unit 23 updates the inference line 56 by recalculating it. Thus, the inference result of the ridge direction made by the inference unit 23 is updated. Additionally, the inference line 56 calculated when the combine harvester 1 reaches the third position P3 is used as the third inference line 56c.

[0136] In this case, the path generation unit 24 updates the target driving path LI by regenerating the target driving path LI based on the third inference line 56c. However, as Figure 11 As shown, at this time, the target driving path LI remains unchanged as the second path LI2. This is because the extension direction of the third inference line 56c is consistent with the extension direction of the second inference line 56b, and the third inference line 56c is located on the extension line 57 of the second inference line 56b.

[0137] Therefore, the combine harvester 1, via the third position P3, automatically performs harvesting along the second path LI2.

[0138] If the structure described above is used, the inference unit 23 infers the ridge direction based on the ridge information of the portion of the ridge located ahead of the combine harvester 1 in the ridge field FI in the direction of travel. Furthermore, it generates a target travel path LI based on the inferred ridge direction. Thus, a travel path management system A capable of generating a suitable target travel path LI in the ridge field FI can be realized.

[0139] [First Other Implementation]

[0140] In the above embodiment, the inference unit 23 determines the target region 53. Furthermore, the inference unit 23 infers the ridge direction by calculating the direction in which the target region 53 extends in the captured image.

[0141] However, the present invention is not limited thereto. Hereinafter, a first other embodiment of the present invention will be described, focusing on the differences from the embodiments described above. The structures other than those described below are the same as those in the embodiments described above. Furthermore, the same reference numerals are used to denote structures identical to those in the embodiments described above.

[0142] In a first alternative embodiment, the target region 53 is not determined. The inference unit 23 in the first alternative embodiment calculates the various directions in which the plurality of ridges 31 extend, or the various directions in which the plurality of furrows 32 extend, or the various directions in which one or more ridges 31 and one or more furrows 32 extend, based on the captured image obtained by the capturing device 40, and infers the ridge direction by calculating the average value of the calculated directions.

[0143] For example in Figure 12 In the example shown, the inference unit 23 calculates an approximate straight line 55 for each of the three first regions 51 and two second regions 52 existing in the analysis object region 50.

[0144] Furthermore, the inference unit 23 calculates five inference lines 56 based on the five approximate straight lines 55. Thus, the inference unit 23 calculates five inference lines 56 corresponding to each of the three first regions 51 and two second regions 52 existing in the analysis object region 50.

[0145] Furthermore, the extension direction of the inference line 56 corresponding to the first region 51 is equivalent to the extension direction of the ridge portion 31 corresponding to the first region 51. Additionally, the extension direction of the inference line 56 corresponding to the second region 52 is equivalent to the extension direction of the groove portion 32 corresponding to the second region 52.

[0146] In other words, the calculation of the inference line 56 corresponding to the first region 51 is equivalent to the calculation of the extension direction of the ridge 31 corresponding to the first region 51. Furthermore, the calculation of the inference line 56 corresponding to the second region 52 is equivalent to the calculation of the extension direction of the groove 32 corresponding to the second region 52.

[0147] exist Figure 12In the example shown, the five calculated inference lines 56 extend in different directions. Furthermore, in this example, the inference unit 23 calculates the average ridge direction 59 by averaging the extension directions of the five calculated inference lines 56. The average ridge direction 59 is equivalent to the "ridge direction" of this invention. In other words, calculating the average ridge direction 59 is equivalent to inferring the ridge direction.

[0148] Thus, in Figure 12 In the example shown, the inference unit 23 calculates the directions in which the three ridges 31 and the two furrows 32 extend based on the captured images obtained by the imaging device 40, and infers the ridge direction by calculating the average value of each calculated direction.

[0149] However, the present invention is not limited thereto. For example, the inference unit 23 may calculate only the directions in which the three ridge portions 31 extend, and simultaneously calculate the average value of the calculated directions to infer the ridge direction. Alternatively, for example, the inference unit 23 may calculate only the directions in which the two furrow portions 32 extend, and simultaneously calculate the average value of the calculated directions to infer the ridge direction.

[0150] [Other Implementation Methods]

[0151] (1) The inference unit 23 can also be configured to reduce the analysis object region 50 after determining the target region 53, thereby increasing the proportion of the target region 53 in the analysis object region 50. In this case, for example, as Figures 6 to 9 As shown in the example, after determining the target region 53, such as Figure 13 As shown, the inference unit 23 can also narrow down the analysis object area 50. In Figure 13 In the example shown, the left portion of the analysis object region 50 is cut off, thereby reducing the size of the analysis object region 50. As a result, the proportion of the target region 53 in the analysis object region 50 increases.

[0152] (2) Alternatively, before the combine harvester 1 begins automatic travel, a reference for the travel direction of the combine harvester 1, i.e., a reference orientation, is set, and a target travel path LI is generated based on the reference orientation. In this case, the reference orientation or the target travel path LI can be updated (corrected) based on the ridge direction inferred by the inference unit 23.

[0153] (3) The path generation unit 24 may also be configured to generate a target travel path LI, so that the combine harvester 1 performs harvesting travel with the left track 11L grounded to the groove 32 corresponding to the target area 53. In this case, the left track 11L is equivalent to the "travel device" of the present invention.

[0154] (4) As an alternative to the left and right tracks 11, multiple wheels may also be provided. In this case, the wheels are equivalent to the "driving device" of the present invention.

[0155] (5) The combine harvester 1 may also be configured not to drive automatically. In this case, the target driving path LI generated by the path generation unit 24 may also be used as guidance for manual driving.

[0156] (6) Some or all of the vehicle position calculation unit 21, driving control unit 22, inference unit 23 and path generation unit 24 may also be located outside the combine harvester 1, for example, they may be located on a management server installed outside the combine harvester 1.

[0157] (7) In the above embodiment, the grid information is the captured image obtained by the imaging device 40. However, the present invention is not limited to this, and the grid information may also be, for example, point group data representing the position and height of an object obtained by LiDAR (Light Detection and Ranging). In this case, LiDAR corresponds to the "acquisition unit" of the present invention.

[0158] (8) When the combine harvester 1 is harvesting in the ridge FI, the imaging device 40 may be configured to acquire only one image. That is, the imaging device 40 may be configured to acquire images (ridge information) over time. Similarly, the inference unit 23 may not be configured to update the inference result of the ridge direction over time. Similarly, the path generation unit 24 may not be configured to update the target travel path LI over time.

[0159] (9) The inference unit 23 can be configured to divide the object region 50 into a first region 51 and a second region 52 by image processing using a neural network that has undergone machine learning.

[0160] (10) It may also be configured as a driving route management program that implements the functions of each component in the above embodiments via a computer. Alternatively, it may be configured as a recording medium that records the driving route management program that implements the functions of each component in the above embodiments via a computer. Alternatively, it may be configured as a driving route management method that executes the operations performed by each component through one or more steps in the above embodiments.

[0161] Furthermore, the structures disclosed in the above embodiments (including other embodiments, hereinafter the same) can be combined and applied in combination with structures disclosed in other embodiments, provided they do not conflict. Additionally, the embodiments disclosed in this specification are illustrative, and the embodiments of the present invention are not limited thereto; appropriate modifications can be made without departing from the purpose of the present invention.

[0162] Industrial availability

[0163] This invention can be used not only for combine harvesters, but also for various work vehicles that operate in ridge fields, such as potato harvesters, carrot harvesters, onion harvesters, onion diggers, and self-propelled management machines.

[0164] Explanation of reference numerals in the attached figures

[0165] 1. Combine harvester (operating vehicle)

[0166] 11R Right Track (Running Device)

[0167] 23. Inference Department

[0168] 24 Path Generation Department

[0169] 31 ridge part

[0170] 32. Groove-like part

[0171] 40. Camera (acquisition unit)

[0172] 50 Analysis Object Area

[0173] 51 First District

[0174] 52 Second Region

[0175] 53 Target Area

[0176] A driving route management system

[0177] FI ridge land

[0178] LI Target Driving Route

Claims

1. A travel path management system for a work vehicle that travels for work in a ridge field, characterized by comprising: a plurality of ridge portions each of which is composed of piled-up soil; and a plurality of furrow portions each of which is provided between two of the ridge portions that are adjacent to each other, the ridge field; a obtaining unit that obtains ridge information related to at least one of the ridge portions and the furrow portions; the obtaining unit configured to obtain the ridge information of a portion of the ridge field that is located ahead of a travel direction of the work vehicle; a inference unit that infers a ridge direction of the ridge portions based on the ridge information obtained by the obtaining unit; a path generating unit that generates a target travel path of the work vehicle based on the ridge direction inferred by the inference unit; a vehicle position calculating unit that calculates position coordinates of the work vehicle over time based on positioning data output from a satellite positioning module possessed by the work vehicle; and a travel control unit that causes the work vehicle to travel automatically along the target travel path based on the position coordinates of the work vehicle received from the vehicle position calculating unit and information indicating the target travel path received from the path generating unit.

2. The travel path management system according to claim 1, characterized in that: the obtaining unit obtains the ridge information over time; the inference unit updates an inference result of the ridge direction over time based on the ridge information obtained by the obtaining unit; and the path generating unit updates the target travel path over time based on the inference result updated by the inference unit.

3. The travel path management system according to claim 1 or 2, characterized in that: the obtaining unit is an imaging device that images a region of the ridge field that is located ahead of the travel direction of the work vehicle; the ridge information is an imaged image obtained by the imaging device; and the inference unit infers the ridge direction based on color information included in the imaged image.

4. The travel path management system according to claim 3, characterized in that: the inference unit divides an analysis target region within the imaged image into a first region corresponding to the ridge portions and a second region corresponding to the furrow portions based on the color information.

5. The travel path management system according to claim 1 or 2, characterized in that: the inference unit infers the ridge direction based on the ridge information obtained by the obtaining unit by calculating an average value of a plurality of directions in which the ridge portions extend, or a plurality of directions in which the furrow portions extend, or a plurality of directions in which one or more of the ridge portions and one or more of the furrow portions extend, simultaneously with the calculation of each of the directions.

6. The travel path management system according to claim 4, characterized in that: the inference unit infers the ridge direction by calculating a direction in which the second region extends in the imaged image. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ The path generating section generates the target travel path such that the work vehicle travels along an extension direction of the trench-shaped section corresponding to the target region, and the work vehicle travels with the travel device of the work vehicle in a state of grounding against the trench-shaped section corresponding to the target region.

7. The travel path management system according to claim 6, wherein In a case where a plurality of the second regions exist in the analysis target region, the inference section determines a second region having a largest area among the plurality of second regions as a target region, and calculates a direction in which the target region extends in the captured image, thereby inferring the ridge direction, The path generating section generates the target travel path such that the work vehicle travels along an extension direction of the trench-shaped section corresponding to the target region, and the work vehicle travels with the travel device of the work vehicle in a state of grounding against the trench-shaped section corresponding to the target region.

8. The travel path management system according to claim 7, wherein The acquisition section acquires the ridge information over time, The inference section updates an inference result of the ridge direction over time, based on the ridge information acquired by the acquisition section, The path generating section updates the target travel path over time, based on the inference result updated by the inference section, The inference section reduces the analysis target region after determining the target region, such that a ratio of the target region in the analysis target region increases.

9. A travel path management program for a work vehicle that travels for work in a ridge field having a plurality of ridge sections composed of piled-up soil and trench-shaped sections provided between two of the ridge sections adjacent to each other, characterized by an acquisition function of acquiring, by a computer, ridge information related to at least one of the ridge sections and the trench-shaped sections, the acquisition function is configured to acquire the ridge information of a portion of the ridge field located ahead of a traveling direction of the work vehicle, an inference function, a path generation function, a self vehicle position calculation function, and a travel control function are implemented by a computer, the inference function infers a ridge direction of an extension direction of the ridge section based on the ridge information acquired by the acquisition function, the path generation function generates a target travel path of the work vehicle based on the ridge direction inferred by the inference function, the self vehicle position calculation function calculates position coordinates of the work vehicle over time based on positioning data output from a satellite positioning module possessed by the work vehicle, the travel control function causes the work vehicle to travel automatically along the target travel path based on the position coordinates of the work vehicle received from the self vehicle position calculation function and information indicating the target travel path received from the path generation function.

10. A recording medium that records a travel path management program for a work vehicle that travels for work in a ridge field having a plurality of ridge sections composed of piled-up soil and trench-shaped sections provided between two of the ridge sections adjacent to each other, characterized by a function of acquiring, by a computer, information about at least one of the ridge and the furrow, the function of acquiring is configured to acquire the ridge information of a portion of the ridge field located ahead of a traveling direction of the work vehicle, a function of inferring, a function of generating a path, a function of calculating a position of the work vehicle, and a function of controlling travel are implemented by a computer, the function of inferring infers a direction of extension of the ridge, i.e., a ridge direction, based on the ridge information acquired by the function of acquiring, the function of generating a path generates a target travel path of the work vehicle based on the ridge direction inferred by the function of inferring, the function of calculating a position of the work vehicle calculates position coordinates of the work vehicle based on positioning data output from a satellite positioning module possessed by the work vehicle, the function of controlling travel causes the work vehicle to perform automatic travel along the target travel path based on the position coordinates of the work vehicle received from the function of calculating a position of the work vehicle and information indicating the target travel path received from the function of generating a path.

11. A travel path management method for a work vehicle that travels for work in a ridge field having a plurality of ridge portions composed of piled-up soil and a furrow portion provided between two ridge portions adjacent to each other, characterized by a step of acquiring information about at least one of the ridge and the furrow, in the step of acquiring, the ridge information of a portion of the ridge field located ahead of a traveling direction of the work vehicle is acquired, the travel path management method includes a step of inferring, a step of generating a path, a step of calculating a position of the work vehicle, and a step of controlling travel, the step of inferring infers a direction of extension of the ridge, i.e., a ridge direction, based on the ridge information acquired by the step of acquiring, the step of generating a path generates a target travel path of the work vehicle based on the ridge direction inferred by the step of inferring, the step of calculating a position of the work vehicle calculates position coordinates of the work vehicle based on positioning data output from a satellite positioning module possessed by the work vehicle, the step of controlling travel causes the work vehicle to perform automatic travel along the target travel path based on the position coordinates of the work vehicle received from the step of calculating a position of the work vehicle and information indicating the target travel path received from the step of generating a path.

Images