Work vehicle

Abstract

To solve the problem that when crops are stored in a container using a harvesting work machine, there is a case where the work is performed while the container is lowered in a farm field, but since a work machine for recovering the container automatically is not available, the recovery work is performed by a human operator who manually runs a transport vehicle.SOLUTION: There is provided a work vehicle in which a robot arm that grabs containers from a farm field catches the containers with a configuration to perform expansion and contraction, and movement in a horizontal direction and a vertical direction with respect to the ground. The containers to be recovered are detected using an imaging device, and running control is performed so that a hand part matches the handle position of the container.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to an unmanned work vehicle capable of automatic travel and automatic loading, which collects containers containing harvested crops in a farm field. [Background technology]

[0002] When crops are removed from the field and stored in containers, the containers are large and heavy, so the work must be done while lowering them into the field, necessitating the use of work vehicles that can retrieve the stored containers unmanned. (Patent Document 1) [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2020-121848 Summary of the Invention [Problem to be solved by the invention]

[0004] Conventional technology uses a gate-shaped frame to move crop storage bags between them. However, the shape of the storage bags is easily deformed, making it difficult to reliably collect them automatically using an imaging device. In particular, it is difficult to automatically detect the hanging part and align the hanging device with that part.

[0005] To provide an unmanned work vehicle that serves as a device for recovering a container having a fixed shape, that can confirm from a distance that a crop is contained in the container, and that travels along a recovery work route that does not involve unnecessary travel of the work vehicle. [Means for solving the problem]

[0006] The first aspect of the present invention is achieved by the following technical means.

[0007] This vehicle has the function of automatically collecting containers filled with harvested produce, and is equipped with an imaging device located at the center line on the left and right sides of the vehicle, which captures images in two directions: the vehicle's travel and the container's position. A virtual line K1 passing through the center position C1 of the left and right hands 55A and 65A that capture the container overlaps and matches a virtual line K3 drawn horizontally from the center position C2 of the container in the image captured by the imaging device, and the center positions C1 and C2. A virtual line K4 drawn vertically from the center position C2 of the container in the image captured by the imaging device matches the virtual line K2 that corresponds to the center line of the vehicle in the fore-and-aft direction. The imaging device is positioned so that the virtual lines K1 and K2 and the center position C1 match, and the virtual lines K3 and K4 and the center position C2 match, even when the imaging device is switched between capturing images horizontally with respect to the ground and capturing images vertically with respect to the ground.

[0008] The second invention is solved by the following technical means.

[0009] The container has a thermolabel (registered trademark) attached to the outside as an identification symbol. Depending on whether or not the container contains crops, a temperature difference is generated due to the amount of light transmitted from the inside, and the identifying characters, shape, image, etc. change. An imaging device analyzes the identifying characters, shape, and image, and a decision is made as to whether the container should be collected. [Effects of the Invention]

[0010] The first invention makes it possible to recognize images of both the running and container catching operations using a single imaging device.

[0011] The second invention makes it possible to automatically recognize whether or not there are crops in the container and whether or not they can be collected, even if the container is located far away, by using an imaging device. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is an overall perspective view of a work vehicle according to the present invention, seen from the left front. [Figure 2] 1 is an overall perspective view of a work vehicle according to the present invention, seen from the left rear. [Figure 3] FIG. 2 is a left side view of the work vehicle of the present invention. [Figure 4] FIG. 1 is a top view of a work vehicle according to the present invention. [Figure 5] FIG. 2 is an overall perspective view of the work vehicle of the present invention, seen from the left front, when the vehicle is traveling with the imaging device facing horizontally to the ground. [Figure 6] FIG. 10 is an overall perspective view of the work vehicle of the present invention from the left front, showing the image capture device being positioned perpendicular to the ground and catching a container. DETAILED DESCRIPTION OF THE INVENTION

[0013] The present invention will be described below with reference to the embodiments shown in the drawings.

[0014] The work vehicle shown in FIGS. 1 to 6 shows an example of this embodiment.

[0015] The background of the work vehicle of the present invention will be explained.

[0016] The work involves pulling crops from the soil in the field, cutting off roots, stems, and leaves as needed, and placing the fruit in a container. However, small-sized vehicles cannot carry a large number of containers filled with harvested crops due to their small size. Therefore, depending on the size of the vehicle, the work may continue by lowering containers into the field as they fill. During this work, containers filled with crops are scattered throughout the field. This requires workers to retrieve the containers. Therefore, if containers are lowered in the center of the field and then retrieved later, it becomes extremely inefficient. Therefore, considering the efficiency of the work, regardless of the amount of crops in the containers, they are currently replaced in a position close to the road surface, where retrieval is easier.

[0017] However, this method results in variations in the amount of containers that can be filled, and requires many containers. Although the weight per load is reduced, the number of operations increases, making transportation from the field using trucks inefficient.

[0018] Given this current situation, it is considered more efficient to use containers until they are full and then lower them into the field when they are full, taking into account subsequent work. When using this work system, it is necessary to retrieve the containers that have been lowered near the center of the field. The present invention is concerned with the task of retrieving containers containing harvested crops, and requires the configuration of a robotic work machine with an automatic driving function.

[0019] The configuration of the work vehicle of the present invention will be described with reference to FIGS. 1, 2, 3 and 4. FIG.

[0020] Wheel 13 is attached to the end of link 11 connected to left frame 10, and wheel 14 is attached to the end of link 12, and links 11 and 12 rotate relative to the left frame to prevent vibration. A reinforced coil spring is attached to the rotation axis at the connecting part to reduce shaking in the rotational direction.

[0021] A wheel 23 is provided at the end of a link 21 connected to the right frame 20, and a wheel 24 is provided at the end of a link 22, and the vibration response is similar.

[0022] To drive the wheels, the wheels 13 are equipped with in-wheel motors 15, which are connected to a steering mechanism 17 (not shown), allowing the vehicle to move forward and backward, change speed by changing the rotation speed of the in-wheel motors, stop, and steer left and right. Each wheel is equipped with a device with a similar configuration.

[0023] Wheel 14 is equipped with a wheel-in motor 16, which is connected to a steering mechanism 18, although not shown. Wheel 23 is equipped with a wheel-in motor 25, which is connected to a steering mechanism 27, although not shown. Wheel 24 is equipped with a wheel-in motor 26, which is connected to a steering mechanism 28, although not shown. The steering mechanism is not essential, and steering can also be achieved by changing the rotation speed of each wheel-in motor. The wheel-in motors are powered by a battery 19. Near battery 19 is a battery management system, although not shown, which receives commands from the management system that controls the operation of the work vehicle and supplies the appropriate amount of power at the appropriate time.

[0024] The working mechanism 30 is placed above this traveling section and has a left side surface 31 and a right side surface 33, which are connected by a frame 35 to form a gate-shaped frame. To enable the robot arm 50 to move up and down within this working mechanism 30, the left side surface 31 is provided with a rail 33 and the right side surface 33 with a rail 34, which serve as rails for the up and down movement of the robot arm 50. The up and down movement of the robot arm 50 is achieved by extending and retracting a central cylinder 36.

[0025] The robot arm 50 has an extension and retraction mechanism to catch a container 100 containing produce. The horizontal arm 62 is made up of a stack of cylinders in several stages, and is shown extended to the shape of horizontal arm 62A. Each arm has an electric cylinder inside, and the extension and retraction of the arm operates by the electric cylinder.

[0026] The position of container 100 in the field is confirmed by imaging device 56 and imaging device 66, and horizontal arm 52 moves to horizontal arm 52A, and horizontal arm 62 moves to horizontal arm 62A. Hand 55 and hand 65 move to hand 55A and hand 65A by extending each arm, and vertical arm 51 and vertical arm 61 reach the positions of vertical arm 51A and vertical arm 61A by being controlled by motor 54A and motor 64A to change the attachment angle and arm dimensions at the joint position of vertical arm 51, vertical arm 61, and horizontal arm 52, horizontal arm 62. This arm control allows hand 55A and hand 65A to move and catch container 100.

[0027] A feature of the weight balance of the work vehicle 1 is that the cylinders 36 for raising and lowering the robot arm 50 are provided above the rear wheels 14 and 24. In the present invention, the center of gravity moves forward as the robot arm extends and retracts. This is addressed by positioning the cylinders 36 above the rear wheels, which are the rearmost ends, when the vehicle is unladen. The battery 19, which serves as the power source for the wheel-in motor, is located below the platform 32 in a thin shape, and serves to position the center of gravity downward.

[0028] Furthermore, when a container is placed on the platform 32, the center of gravity is stabilized at the bottom in the center, ensuring excellent balance during subsequent work. The platform has protrusions in some areas to prevent the loaded container from easily shifting position due to vibration.

[0029] A feature of the hand section of the work vehicle 1 is that the tips of hands 55 and 65 are tapered in the vertical direction and have an arc-shaped curve in the front-to-back direction, making it easy to fit into the handle of a container. Therefore, the work vehicle 1 can approach a container placed in a field so that it faces the front of the machine, and once it is ready to store the container, it can operate the robot arm 50, or it can operate the robot arm 50 while traveling.

[0030] The rails 33 and 34, which also serve as the mast at the rear of the vehicle, have a wider pitch than the lower platform 32 and are almost the same width as the running wheels, so the width of the container can be fully utilized and the width of the vehicle can be made small. Depending on the configuration, only the top plate of the platform can be made wider by overlapping with the running wheels when viewed from above.

[0031] A satellite positioning unit 90 and an inertial measurement unit 91 are arranged on the frame 35, which is the uppermost position of the work mechanism 30, to enable automatic driving of the work vehicle 1. Although not shown, Lidar is arranged near the satellite positioning unit 90 and the inertial measurement unit 91 to provide an overview of the positional relationship between the work vehicle 1 and the field, eliminating any difference in position with the satellite positioning unit 90 and creating conditions that make it easier to calculate the relationship between the imaging positions. Furthermore, if an ultrasonic or millimeter-wave radar device is arranged as a distance measurement device other than Lidar, it can also be used as an obstacle sensor.

[0032] Since there are imaging devices 56 and 57 on the left and right and they function as stereo cameras, it is possible to calculate three-dimensional dimensions, and the dimensional and angular deviations in both the horizontal and vertical directions from the ground become clear, and this distance deviation and angle are controlled to match by steering the work vehicle and the extension and contraction amount of the robot arm 50.

[0033] To further improve accuracy, an imaging device 70 is installed on the center line of the robot arm 50, and this imaging device is used to capture images for navigation. When the robot arm approaches within a predetermined distance, the imaging device is turned toward the ground to capture images of the inside of the container.

[0034] The more imaging devices there are, the better the accuracy, but in the first invention, it is possible to detect containers, control their travel, check the contents of the containers, and control their alignment with the container center using only the imaging devices 70 on the center line of the robot arm 50.

[0035] The first aspect of the invention will be described with reference to FIGS.

[0036] No components that obstruct imaging are placed in the imaging direction of imaging device 70 or imaging device 70A. This position is the left-right center position of the work vehicle, and is also the center position of hands 55, 55A and hands 65, 65A. In other words, the center of the captured image is the same as the center position of the traveling vehicle, and if the center position of the image matches the target center position for traveling, it can be seen that the vehicle can travel to a position where hands 55, 55A and hands 65, 65A can capture the image uniformly.

[0037] Therefore, when detecting a container to be collected from a distance and driving, the vehicle is positioned facing forward, which is the position of the imaging device 70, and the center position of the image of the imaging device is aligned with the center of the target object.

[0038] Figure 5 shows the vehicle traveling forward, aiming at the container 100 to be retrieved, where the imaging device 70 is located. In this figure, imaging devices 56 and 66 are installed to further improve image accuracy, and a stereo camera function is also used. The installation of imaging devices 56 and 66 is used to correct the left and right tilt of the work vehicle. Since tilt affects the catching of the container, the work vehicle must be kept level, and in this configuration, this is addressed by adjusting links 11, 12, 21, and 22 to match the captured image.

[0039] Tilt correction can also be performed by the imaging device 70. This involves analyzing one image, but it is possible to correct tilt by analyzing whether the horizontal reference position of the target position is horizontal in the captured image.

[0040] 5 shows that the image capture device 70 and the image capture devices 56 and 66 are installed simultaneously, but if either one is installed, it is possible to adjust for the tilt of the vehicle body. Furthermore, during this analysis, accuracy can be further improved by using data from the inertial measurement unit 91 and making corrections that take into account the unevenness of the field in front of and behind the vehicle body.

[0041] 6 shows the state when catching a container, and the camera moves downward to the position of the image capturing device 70A. This image capturing position is a position on a horizontal plane with the ground, which is the center position C1 of the arms 51A and 61A, and when an imaginary line K1 connecting the center positions C1 of the hands 55A and 65A coincides with an imaginary line K3 drawn horizontally from the center position C2 of the container in the image captured by the image capturing device 70A to the short side 104 or long side 105 of the container, the camera is controlled so that the hands 55A and 65A can catch the handle of the container 100.

[0042] To align the container, each wheel-in motor is operated to rotate each wheel individually, thereby aligning the vehicle body.

[0043] In the first aspect of the present invention, in order to control traveling and catch a container by using an imaging device 70 that captures images in two directions, that is, traveling and the container position, located at the center line position of the left and right sides of the vehicle, the imaginary line K1 corresponding to the width distance passing through the center position C1 of the left and right hands 55A and 65A is made to overlap with the center position C1 and center position C2 of an imaginary line K3 drawn horizontally from the center position C2 of the container in the image of the container captured by the imaging device 70, and they are made to coincide simultaneously on a plane, and the imaginary line K4 drawn vertically from the center position C2 of the container in the image of the container captured by the imaging device 70 coincides with the imaginary line K2 corresponding to the center line in the fore-and-aft direction of the vehicle, and even when the imaging device 70A is switched between capturing images horizontally with the ground and capturing images perpendicular to the ground, when switched to capturing images for the container position perpendicular to the ground, the center position of the imaging device is positioned at a position that coincides with the center position C1 on the center line position of the left and right sides of the vehicle that satisfies the above-mentioned matching condition.

[0044] In other words, the position where virtual line K1, virtual line K2, and center position C1, and virtual line K3, virtual line K4, and center position C2 coincide is a position where the container handle can be caught, there is no need to make driving corrections, and the container and vehicle are facing straight ahead, but the direction of travel (the direction in which the ridges or furrows extend) and the direction of the vehicle body may be different, and the positional relationship of the corners, etc. is stored by the imaging device as driving data.

[0045] To further explain this, consider the case where separate imaging devices are installed, one positioned horizontally relative to the ground and the other positioned vertically relative to the ground, and each captures an image. This approach requires that each imaging device be equipped with an absolute position reference, and position reference data be entered into the captured images, correcting the position reference for each image before comparing them. The position data used in this process comes from the satellite positioning device 90. Because this satellite positioning device 90 calculates the direction of travel by calculating the direction and distance traveled, if the robot performs a movement more like a rotation than a steering operation, such as when catching the handle of a container in this control, the measurements from the satellite positioning device 90 and inertial measurement unit 91 will fall within the error data range due to minute changes. Therefore, the position in the next process will be determined to have been no movement, without taking into account the rotational movement of the catch control. This increases the risk of misdirection and collision with a ridge or furrow.

[0046] A feature of the present invention is that the orientation and direction of the work vehicle and the orientation and direction of the container are calculated using values ​​calculated from the satellite positioning unit 90, inertial measurement unit 91, and imaging device data when the work vehicle travels long distances. The data reference from the imaging device used for travel can be used as the reference for the container position, so the orientation and direction of the container are accurately detected. Therefore, even when the work vehicle is rotated to match the container using the imaging device, the difference from the direction and orientation it has traveled up to that point is clear, so it is possible to accurately start traveling even after catching the container.

[0047] However, if images of individual containers are taken, the image data is calculated based only on displacement from a reference orientation and direction, and therefore only shows relative position relationships. Data from the satellite positioning device 90, which is an absolute value, does not respond to minute rotations, so control must prioritize the image capture device. In this case, although it is possible to catch the container, subsequent travel must be controlled by the image capture device, resulting in low travel accuracy immediately after the start of operation.

[0048] Since this proposal is primarily intended for use in harvesting vegetables, etc., the fields have ridges and furrows, making it impossible to perform control such as making corrections while traveling. Due to these conditions, the image data horizontal to the ground and the orientation and direction of the work vehicle must match, the image data vertical to the ground and the orientation and direction of the hand must match, and by rotating the image data horizontal to the ground and the image data vertical to the ground along the traveling direction axis using the same imaging device, the virtual lines K1 and K2 and the center position C1, and the virtual lines K3 and K4 and the center position C2, must match. By installing the imaging device in this position, the conditions can be met, making it possible to accurately detect the differences in orientation and direction from the traveling direction that occur when rotating to catch a container.

[0049] The reason why the travel control and container catch control can be linked in this way by rotating and moving a single imaging device is that, as mentioned above, imaging device 70 is positioned so that virtual lines K1, K2, and center position C1 coincide with each other, and virtual lines K3, K4, and center position C2 coincide with each other.

[0050] 6, the positional relationship of the container images captured by the imaging device is expressed as a virtual line at the center position in the height direction of the container 100. In reality, virtual lines K1, K2, and center C1 represent the positional relationship based on the positions of the hand unit and vehicle from the currently captured image, and virtual lines K3, K4, and center C2 represent the positional relationship based on the position of the container from the currently captured image, and are image processing of the same image, and both occur during the image analysis process in the imaging device or the calculation unit. However, for the sake of visual representation, in FIG. 5, the positional relationship based on the positions of the hand unit and vehicle is represented at the position where hands 55A and 65A pass, and the positional relationship based on the position of the container is represented by a virtual line at the center position in the height direction of the container 100.

[0051] The basic image of the imaging device shows the vertical and horizontal relationship of the maximum range that can be captured by the imaging device. By arranging the imaging device in the above-mentioned position, the image position becomes the center image position of the vehicle, and the direction of the captured image is arranged to correspond to the movement of the vehicle itself, and can be used for driving control.

[0052] At the position where the imaging device 70 is rotated to the imaging device 70A, the virtual lines K1 and K2 of the hand part and the center position C1 are aligned with the horizontal and vertical lines of the maximum range captured by the imaging device 70A and the center position of the image, and thus the positional relationship of the hand part, which is the operation of traveling and catching the container, becomes continuous image data. Therefore, by creating images based on the hand and the container using only the images currently captured within the imaging device and calculating the deviation angle and deviation position, it is possible to immediately match the positional deviation by directly controlling the difference distance as the amount of movement of the wheels.

[0053] It is also possible to use a technology in which an image of a container placed in a normal state is taken as a base image, and the long and short sides of the container are compared.Even when comparing images in this way, matching is not possible unless the relationship between the reference image and the vehicle matches.Even if images are switched between when the vehicle is moving and when the container is caught, the same conditions are met, which makes it possible to speed up calculations, improve accuracy, and reduce costs.

[0054] When aligning with a container, even if the vehicle's position is controlled by the rotation of the wheels, unevenness in the field can cause the container or vehicle body to twist, making it impossible to catch the container. In this case, it is necessary to adjust the position and angle of arms 51 and 61. If arms 51 and 61 or motors 54 and 64 are captured in the image of the imaging device due to their controlled movements, this can cause erroneous recognition in image processing. For this reason, motors 54, 54A and 64, 64A, which adjust the position and angle of each arm, are located at the ends of the extendable horizontal arms 52A and 62A, respectively, so that they do not appear in the imaging device during image analysis.

[0055] The switching of the imaging device 70 between imaging in a direction horizontal to the ground and imaging in a direction perpendicular to the ground will now be described.

[0056] When the imaging device 70 faces downward to the position of the imaging device 7A, which is the position when the imaging device 70 faces downward to take an image of the inside of the container, there is an advantage that adhesion of dust to the lens can be suppressed.

[0057] The position of the imaging device 70 changes significantly between a position where it captures images in front and a position where it captures images facing directly downward, but because the imaging range is wide, each image is captured widely, and although there is some overlapping in the image data, these areas are combined to create the image. Because this image processing technology is used, even when the imaging device is changed from a position facing forward to a position facing directly downward, the transition can be carried out smoothly.

[0058] In controlling the travel switching, once the position and direction of the container are identified, the camera moves from the imaging device 70 to the imaging device 70A. Note that even if the imaging device is pointed directly downward, it is necessary to travel to a point where the container cannot be seen in its entirety, and this position is the predetermined distance for switching.

[0059] The following describes the state where the image capture device 70 is positioned facing the direction of travel. The container 100 is designed to recognize the state of crops contained in the container from a distance. This recognition is achieved by dividing the recognition image into sections on the periphery of the container.

[0060] The images recognized by the imaging device can be classified using characters, shapes, image identification symbols, etc. that can be recognized by the imaging device. By classifying the images using the imaging device in this way, it becomes possible to read the status of the container from a distance and determine whether the container should be recovered. This method of identifying containers will be explained later.

[0061] Once the container to be collected is identified and its location is detected, a travel route is created. The travel route is determined based on the data registered on the map using a satellite positioning device 90 and an inertial measurement unit 91.

[0062] The imaging device is used to check whether the vehicle is traveling correctly along the route, to check for obstacles, to check the position of ridges, and to check for unevenness in the field.

[0063] The imaging devices shown in Figures 1, 5, and 6 are of a stereo type that can measure distance from images, and are also capable of estimating and determining three-dimensional distance by combining the differences between left and right images.

[0064] A recognition area for identification is provided on the outer surface of the container. The recognition area has areas where light passes and areas where it does not; when there are no crops inside the container, light passes through, and when there are crops inside, light does not pass through, creating a temperature difference depending on the amount of light. When a Thermo Label (registered trademark) is attached, the color changes depending on the temperature, and this change can be expressed as an image or text, allowing the imaging device to recognize it.

[0065] By changing the way the image is recognized by the imaging device, it becomes possible for the imaging device to determine whether there are crops inside the container even if the work vehicle is far away.

[0066] Labels 102 and 103 on the outer surface are arranged in a grid pattern, which allows light to pass through and block out. The position where these labels are attached needs to be such that the container is shaped to allow light to easily escape, but the use of labels has the advantage that the same containers can be used.

[0067] In the second invention, a container has a thermolabel (registered trademark) attached to the outside as an identification symbol, and a temperature difference is generated depending on whether or not the container contains crops, depending on the amount of light transmitted from the inside, so that the identifying characters, shape, image, etc. change.The identifying characters, shape, and image are analyzed by an imaging device, and a decision is made to retrieve the container.

[0068] In another configuration, when a load is applied to the bottom of the housing, a two-dimensional code on the outer surface lights up using a battery or the like built into the housing.

[0069] While the above configuration is automatic, there is also a manual method. This involves attaching a mask sheet to the label. When the container is full, a person flips over the mask sheet to expose the code displayed on the label. The mask sheet is lowered downwards, eliminating the need for a lock.

[0070] By using a predetermined color for the identification symbol or the container itself, the color can be recognized and distinguished using the RGB color analysis function of the imaging device.

[0071] An imaging device is installed at the top of the frame 35 of the work vehicle 1, and the image obtained by the imaging device is binarized and a typical algorithm used to detect objects such as YOLO (You Only Look Once) is used to perform image segmentation, obtain the difference between the image's center of gravity and the center of the camera's angle of view, and use this as data for position correction.

[0072] The containers will be a different color so as not to be confused with the crops, soil, or residual sand left in the field.

[0073] The image recognition system ignores objects that are smaller than a certain value, and if there are multiple objects, it prioritizes the larger ones and collects them.

[0074] If the size of the container is known in advance, the appropriate height of the hand unit can be determined based on the image size.

[0075] A sensor consisting of a contact switch or proximity switch can be installed at the lower end of the left and right connecting members of the second arm to detect the appropriate height by touching the top end of the container.

[0076] If the furrow is wide enough, one method is to enter forward and then turn around and transport the container after retrieving it. Conversely, if the furrow is narrow, the container can be grabbed and retrieved from the furrow and then moved backward to the headland. It is also possible to enter the furrow at a right angle by pointing the aircraft directly sideways and then submitting it at a right angle.

[0077] During such driving, in addition to the satellite positioning device 90 and inertial measurement unit 91, a LiDAR device (not shown) at the top of the mast is also used to measure positioning using images, and this data is combined with the data from the satellite positioning device 90, allowing the vehicle to travel along a route by referring to the registered driving route in the map data.

[0078] If the travel route in the field is already determined, it is possible to register the route by teaching.

[0079] The field is uneven, and the robot is driven to avoid areas where the values ​​of the inertial measurement unit 91 suddenly rise or fall, as these are areas where there is a risk of the robot falling.

[0080] If the satellite positioning device 90 is unable to estimate its own position or if the imaging device cannot be used due to dust or rain, it will immediately stop.

[0081] When loading onto the bed of a truck or the like, the imaging device 80 is positioned forward and approached to a predetermined distance. Then, the container is raised until the bottom end of the container is higher above ground level than the bed position, and then the container is aligned by moving forward a predetermined distance. [Explanation of symbols]

[0082] 1 Work vehicle 50 Robot Arm 51 Vertical Arm 52 horizontal arm 56 Imaging device 66 Imaging device 70 (70A) Imaging device 100 containers K1 virtual line (hand) K2 virtual line (hand) K3 virtual line (container) K4 virtual line (container) C1 center position (hand) C2 central position (container)

Claims

1. A vehicle having a function of automatically collecting containers containing harvested products, Imaging devices are provided at the center line positions on the left and right sides of the vehicle to capture images in two directions: traveling and container position; The imaginary line (K1) passing through the center position (C1) of the left and right hands (55A) and (65A) that capture the container is A virtual line (K3) drawn horizontally from the center position (C2) of the container in the container image captured by the imaging device is superimposed on the center position (C1) and the center position (C2), and the virtual line (K3) is superimposed on the center position (C1) and the center position (C2). a virtual line (K4) drawn perpendicularly from the center position (C2) of the container in the container image captured by the imaging device coincides with a virtual line (K2) corresponding to the center line of the vehicle in the longitudinal direction; Even if the imaging device switches between imaging for traveling horizontally with the ground and imaging for container positioning vertically with the ground, A work vehicle in which an imaging device is installed at a position where the virtual line (K1), the virtual line (K2), and the center position (C1) coincide with the virtual line (K3), the virtual line (K4), and the center position (C2).

2. A container that has a Thermo Label (registered trademark) on the outside as an identification symbol, Depending on whether or not there are crops inside the container, a temperature difference is generated depending on the amount of light passing through from the inside, and the identifying letters, shapes, images, etc. change. Characters, shapes, and images identified by the imaging device are analyzed, The work vehicle according to claim 1, wherein a decision is made as to whether or not to retrieve the container.

Images