Automatic travelling system and automatic travelling method

Abstract

To provide an automatic travelling system and an automatic travelling method capable of allowing a travelling device to appropriately automatically travel with a simple process along a known passage.SOLUTION: Each of a plurality of markers 30 is arranged on a passage 100 so that normal directions ND of the two-dimensional markers 30 are within a prescribed angle range from a direction in which the passage 100 extends when a travelling device 1 and the plurality of markers 30 are viewed from above. When the travelling device 1 is straightly travelling along the passage 100, an angle formed by a direction of photographing optical axes OF and OB of two-dimensional cameras 5F and 5B arranged on the travelling device 1 and the normal direction ND of the markers 30 arranged on the passage 101 on which the travelling device 1 is straightly travelling is within a range of 25 degrees and 65 degrees.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present disclosure relates to an automatic driving system that automatically drives a traveling device along a known path, and an automatic driving method executed in the automatic driving system. [Background technology]

[0002] There are known techniques for automatically driving a traveling device along a path. For example, an automatic traveling device described in Patent Document 1 detects marks that are attached in advance to locations adjacent to a traveling guide line based on a captured image, and travels along the traveling guide line based on the detection results. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2017-111554 Summary of the Invention [Problem to be solved by the invention]

[0004] The inventors of the present invention have thoroughly investigated a technology for automatically driving a traveling device along a known path with simple processing. First, a method for automatically driving a traveling device could be considered, in which the distance and direction of the traveling device are controlled based solely on map data of the known path. However, in this case, even a slight change in the environment in which the path is installed increases the possibility of inappropriate automatic driving. For example, when driving a traveling device along a path installed in a crop cultivation field, the environment changes with the growth and harvest of the crops, changes in the soil, etc., making appropriate automatic driving difficult using only a map of the path. Therefore, it is desirable to accurately determine the current location of the traveling device.

[0005] However, using GPS or other technologies can easily reduce the accuracy of the current location, especially when the traveling device is automatically traveling within a narrow area. Furthermore, there are places where GPS or other technologies cannot be used, such as inside buildings and greenhouses. Therefore, the inventors of the present invention investigated a method for determining the current location of a traveling device by installing markers at predetermined locations along a passageway and measuring the relative position of the traveling device to the installed markers. It is possible to measure the relative position of the traveling device to the marker by installing a 3D camera (three-dimensional camera) on the traveling device and processing images of the marker captured by the 3D camera. However, 3D cameras are expensive, and processing images captured by a 3D camera requires a significant processing load. Therefore, to minimize the processing load, it is desirable to use a 2D camera. However, it was newly discovered that simply using a marker and a 2D camera reduces the accuracy of measuring the traveling device's relative position to the marker.

[0006] A typical object of the present disclosure is to provide an automatic driving system and an automatic driving method that are capable of automatically driving a traveling device appropriately along a known path with simple processing. [Means for solving the problem]

[0007] An automated driving system provided by a typical embodiment of the present disclosure includes a traveling device that travels along a passageway and a control unit that controls the traveling of the traveling device, and is an automated driving system that automatically drives the traveling device along the known passageway, wherein two-dimensional markers having specific geometric characteristics are arranged at each of a plurality of predetermined positions on the passageway, the traveling device includes a traveling drive unit that is capable of changing the traveling direction of a main body and causes the main body to travel in the traveling direction, and a 2D camera that is provided on the main body and is capable of photographing the markers arranged on the passageway, and the control unit processes an image photographed by the 2D camera to identify at least one of the markers included in the image and acquires the distance and direction of the traveling device relative to the identified marker as a relative position, and and a current position acquisition step of acquiring the current position of the traveling device within the map of the passage based on map data showing the arrangement of the markers and the relative position of the traveling device with respect to the identified markers; and a traveling control step of controlling the driving of the traveling drive unit based on the current position of the traveling device acquired in the current position acquisition step, thereby causing the traveling device to automatically travel along the passage. When the traveling device and the multiple markers are viewed from above, each of the multiple markers is arranged on the passage so that the normal direction of the two-dimensional marker is within a predetermined angle range with respect to the direction in which the passage extends, and when the traveling device is traveling straight along the passage, the angle formed by the direction of the imaging optical axis of the 2D camera provided on the traveling device and the normal direction of the marker arranged on the passage in which the traveling device is traveling straight is within a range of 25 degrees to 65 degrees.

[0008] An automatic traveling method provided by a typical embodiment of the present disclosure is an automatic traveling method executed by an automatic traveling system including a traveling device that travels along a passage and a control unit that controls the traveling of the traveling device, and that automatically travels the traveling device along a known passage, wherein two-dimensional markers having specific geometric characteristics are arranged at each of a plurality of predetermined positions on the passage, the traveling device including a traveling drive unit that is capable of changing a traveling direction of a main body and that travels the main body in the traveling direction, and a 2D camera that is provided on the main body and is capable of photographing the markers arranged on the passage, the automatic traveling method including a relative position acquisition step of processing an image photographed by the 2D camera to identify at least one of the markers included in the image and acquiring a distance and direction of the traveling device relative to the identified marker as a relative position, and a relative position acquisition step of processing a plurality of markers in a map of the passage. The method includes a current position acquisition step of acquiring the current position of the traveling device within the map of the passage based on map data indicating the arrangement of each of the markers and the relative position of the traveling device with respect to the identified markers, and a travel control step of automatically traveling the traveling device along the passage by controlling the drive of the travel drive unit based on the current position of the traveling device acquired in the current position acquisition step, wherein when the traveling device and the multiple markers are viewed from above, each of the multiple markers is arranged on the passage so that the normal direction of the two-dimensional marker is within a predetermined angle range with respect to the direction in which the passage extends, and when the traveling device is traveling straight along the passage, the angle formed by the direction of the shooting optical axis of the 2D camera provided on the traveling device and the normal direction of the marker arranged on the passage in which the traveling device is traveling straight is within a range of 25 degrees to 65 degrees.

[0009] According to the automatic driving system and automatic driving method disclosed herein, the traveling device can be automatically driven appropriately along a known path with simple processing. [Brief explanation of the drawings]

[0010] [Figure 1]FIG. 2 is a perspective view of the traveling device 1 as seen from diagonally above on the left. [Figure 2] 2 is a block diagram showing the electrical configuration of the automatic driving system 10. FIG. [Figure 3] 1 is a schematic diagram showing a traveling device 1 automatically traveling along a known path 100 on which markers 30 are installed, as viewed from above. [Figure 4] FIG. 1 is a view of one of the two-dimensional markers 30 as seen from the front (normal direction). [Figure 5] 5 is a schematic diagram of the marker 30 shown in FIG. 4 when viewed from the left diagonally forward. [Figure 6] 1 is a schematic diagram showing a traveling device 1 seen from above immediately before entering a branch passage 102 from a main passage 101. FIG. [Figure 7] 1 is a schematic diagram of a traveling device 1 seen from above when the traveling device 1 starts to enter a branch passage 102 from a main passage 101. FIG. [Figure 8] 1 is a schematic diagram showing the traveling device 1 seen from above after entering the branch passage 102 from the main passage 101. FIG. [Figure 9] 10 is a diagram showing a schematic diagram of the relationship between the normal direction ND and the photographing optical axis O of the 2D camera 5 when the marker 30 is viewed from a horizontal direction perpendicular to the normal direction ND. [Figure 10] 2 is a flowchart of an automatic driving control process executed by a control unit of the automatic driving system 10. DETAILED DESCRIPTION OF THE INVENTION

[0011] <Summary> The autonomous driving system exemplified in the present disclosure includes a traveling device that travels along a passageway and a control unit that controls the traveling of the traveling device, and automatically drives the traveling device along a known passageway. The autonomous driving method of the present disclosure is executed by the autonomous driving system. Two-dimensional markers having specific geometric characteristics are pre-placed at each of a plurality of predetermined positions along the passageway. The traveling device includes a traveling drive unit and a 2D camera (e.g., a monocular 2D camera). The traveling drive unit is capable of changing the traveling direction of the main body of the traveling device and causes the main body to travel in the traveling direction. The 2D camera is provided on the main body of the traveling device and can capture images of the markers placed on the passageway. The control unit executes a relative position acquisition step, a current position acquisition step, and an autonomous driving step. In the relative position acquisition step, the control unit processes an image captured by the 2D camera to identify at least one marker included in the image and acquires the distance and direction of the traveling device relative to the identified marker as a relative position. In the current position acquisition step, the control unit acquires the current position of the traveling device within the map of the passage based on map data indicating the arrangement of each of the multiple markers on the map of the passage and the relative position of the traveling device with respect to the identified marker. In the travel control step, the control unit controls the drive of the travel drive unit based on the current position of the traveling device acquired in the current position acquisition step, thereby causing the traveling device to automatically travel along the passage.

[0012] When the traveling device and the multiple markers are viewed from above, each of the multiple markers is placed on the passage so that the normal direction of the two-dimensional marker is within a predetermined angle range with respect to the direction in which the passage extends. When the traveling device is moving straight along the passage, the angle formed by the direction of the imaging optical axis of the 2D camera attached to the traveling device and the normal direction of the marker placed on the passage in which the traveling device is moving straight is within a range of 25 degrees to 65 degrees.

[0013] According to the autonomous driving system disclosed herein, a 2D camera mounted on the traveling device captures images of markers placed at predetermined locations along a path, thereby accurately acquiring the current position of the traveling device. Based on the acquired current position, the traveling device automatically navigates along the path with high accuracy. This makes the system less susceptible to changes in the environment in which the path is located, and reduces the processing load compared to using a 3D camera. Furthermore, the inventors of the present application have discovered that when the normal direction of a two-dimensional marker is parallel to the optical axis of the 2D camera (e.g., when the 2D camera captures an image of the marker from the front), the measurement accuracy of the 2D camera's relative position (particularly its direction) to the marker is likely to decrease. This is believed to be due to the fact that, when the normal direction of the marker is parallel to the optical axis of the 2D camera, the normal direction of the marker detected based on the two-dimensional image is detected as being exactly opposite to the actual normal direction. In the automated driving system disclosed herein, the 2D camera and markers are arranged so that when the traveling device is traveling straight along a passage, the angle formed by the direction of the imaging optical axis of the traveling device's 2D camera and the normal direction of the marker placed on the passage is within a range of 25 degrees to 65 degrees. As a result, the frequency with which the normal direction of the marker and the imaging optical axis of the 2D camera become parallel decreases. Therefore, the relative position of the 2D camera (traveling device) with respect to the marker can be measured with high accuracy based on the image captured by the 2D camera while suppressing an increase in processing load.

[0014] A specific method for acquiring the relative position of the traveling device with respect to the marker included in the image by processing the image can be selected as appropriate. For example, the control unit may acquire the direction of the traveling device with respect to the marker (e.g., the direction of the marker as seen from the traveling device) based on the position of the marker in the image captured by the 2D camera (e.g., the position of the marker in the coordinate system of the image) and the orientation of the marker in the image (e.g., the shape of the marker). The control unit may also acquire the distance of the traveling device with respect to the marker (the distance between the marker and the traveling device) based on the size of the marker in the image captured by the 2D camera.

[0015] Various types of markers can be used. For example, AR markers, in which a geometric shape is drawn within a rectangular two-dimensional area, may be placed in a passageway. In this case, processing the image captured by the 2D camera makes it easier to properly measure the relative position of the 2D camera with respect to the AR marker. Note that the geometric characteristics attached to each of the multiple markers may be unique to each marker. In this case, the control unit can properly identify which of the multiple markers is the marker appearing in the image captured by the 2D camera.

[0016] The configuration of the control unit can also be selected appropriately. For example, the automated driving system may include a driving device and a control unit (e.g., a personal computer, a server, etc.). The driving device and the control unit may be communicatively connected via wireless communication or the like. The controller of the control unit may function as a control unit that executes a relative position acquisition step, a current position acquisition step, and a driving control step for controlling the automated driving of the driving device. Furthermore, the controller of the control unit and the controller of the driving device may cooperate to execute the relative position acquisition step, the current position acquisition step, and the driving control step. In this case, the driving device can be automatically driven appropriately by at least a part of the control being executed by the controller of the control unit, without improving the processing power of the controller of the driving device. Furthermore, the controller of the driving device may execute the relative position acquisition step, the current position acquisition step, and the driving control step independently. In this case, the automated driving system does not need to include a control unit.

[0017] A plurality of markers may be placed in the passage so that the normal direction of the markers is in the range of 80 degrees to 100 degrees with respect to the direction in which the passage extends. The direction of the imaging optical axis of the 2D camera provided on the traveling device may be tilted in the range of 25 degrees to 65 degrees with respect to the straight direction extending forward and backward from the traveling device.

[0018] In this case, the two-dimensional marker can be easily installed because it only needs to be positioned so that its planar direction is along the direction in which the passage extends (i.e., so that the normal direction of the marker is perpendicular to the center of the passage). Furthermore, unlike when the planar direction of the marker is tilted relative to the direction in which the passage extends, the 2D camera of the traveling device can photograph the marker from a position closer to the marker in the direction of travel, and can also photograph the marker from a position farther from the marker in the direction of travel. This makes it easier to obtain the relative position of the traveling device with higher accuracy.

[0019] In this disclosure, the "straight direction" includes not only the direction in which the traveling device travels straight forward, but also the direction in which the traveling device travels straight backward. In other words, the "straight direction" in this disclosure refers to the forward / backward direction when the traveling device travels straight forward or backward without turning. In addition, if multiple 2D cameras are installed on the traveling device, the current position of the traveling device can be measured with high accuracy as long as the imaging optical axis of at least one 2D camera is tilted within a range of 25 degrees to 65 degrees with respect to the straight direction.

[0020] The direction of the imaging optical axis of one or more 2D cameras provided on the traveling device may be tilted within a range of 25 degrees to 65 degrees to either the left or right with respect to the straight direction extending forward and backward from the traveling device.

[0021] In this case, the 2D camera captures an image of either the left or right marker of the traveling device. According to the autonomous driving system disclosed herein, the current position of the traveling device is acquired based on the distance and direction (relative position) of the traveling device relative to the marker. Therefore, even if the 2D camera does not capture images of both the left and right markers of the traveling device, the current position of the traveling device can be appropriately acquired. Furthermore, when capturing images of only the left or right side of the traveling device, the number of 2D cameras installed on the traveling device can be reduced compared to when capturing images of both the left and right sides. Therefore, autonomous traveling of the traveling device can be achieved with a simpler configuration and processing.

[0022] Two 2D cameras may be provided on the main body of the traveling device. Here, the shooting direction is either the left or right side with respect to the straight traveling direction extending forward and backward from the traveling device. The direction of the shooting optical axis of one of the 2D cameras may be tilted within a range of 25 degrees to 65 degrees toward the shooting direction with respect to the forward straight traveling direction. The direction of the shooting optical axis of the other 2D camera may be tilted within a range of 25 degrees to 65 degrees toward the shooting direction with respect to the rearward straight traveling direction.

[0023] In this case, by capturing images of only one of the left and right sides of the traveling device, the number of 2D cameras installed on the traveling device and the increase in processing load are appropriately suppressed. Furthermore, by capturing images of the left diagonally forward and left diagonally backward, or the right diagonally forward and right diagonally backward, of the traveling device with a 2D camera, the current position of the traveling device in the direction along the passage can be obtained with higher accuracy. This makes it easier for the traveling device to travel automatically.

[0024] However, the specific installation method of the 2D camera relative to the main body of the traveling device may be changed. For example, when two 2D cameras are installed on the traveling device, the direction of the imaging optical axis of one 2D camera may be tilted to the right with respect to the direction of travel within a range of 25 degrees to 65 degrees, and the direction of the imaging optical axis of the other 2D camera may be tilted to the left with respect to the direction of travel within a range of 25 degrees to 65 degrees. Even in this case, the markers are photographed by the 2D camera, thereby appropriately controlling the automatic traveling of the traveling device. Furthermore, the number of 2D cameras installed on the traveling device may be one or three or more.

[0025] It is also possible to change the installation angle of the marker relative to the passage and the installation angle of the 2D camera relative to the traveling device. For example, the direction of the imaging optical axis of the 2D camera attached to the traveling device may be parallel to the straight direction extending forward and backward from the traveling device, or at 90 degrees. The normal direction of the marker may be tilted within a range of 25 degrees to 65 degrees with respect to the direction in which the passage extends. Even in this case, the frequency with which the normal direction of the marker and the imaging optical axis of the 2D camera become parallel decreases. Therefore, the relative position of the 2D camera (traveling device) relative to the marker can be measured with high accuracy while suppressing an increase in processing load.

[0026] The control unit may further execute a reliability calculation step of calculating the reliability of the relative position acquired based on the markers based on the states of the markers in the images captured by the 2D camera. In this case, the current position of the traveling device is acquired based on the calculated reliability, which further facilitates improving the accuracy of the acquired current position.

[0027] In the reliability calculation step, the control unit may calculate a lower reliability when the shape of the marker captured in the image captured by the 2D camera is closer to the shape captured when the 2D camera's imaging optical axis is aligned with the normal direction of the marker (i.e., when the 2D camera is directly facing the marker).

[0028] As mentioned above, when the normal direction of the two-dimensional marker is parallel to the optical axis of the 2D camera, the measurement accuracy of the relative position of the 2D camera to the marker tends to decrease. Therefore, the closer the shape of the marker in the image is to the shape captured with the 2D camera facing the marker directly, the lower the reliability calculated, making it easier to calculate the reliability appropriately.

[0029] The control unit may calculate a lower reliability when the shape of the marker in the image captured by the 2D camera is closer to a parallelogram than when the shape of the marker is a trapezoid, in the reliability calculation step.

[0030] The inventors of the present application have discovered that the closer the external shape of a marker in a two-dimensional captured image is to a parallelogram, the more likely it is that the accuracy of measuring the relative position of the 2D camera relative to the marker decreases. This is thought to be due to the phenomenon that, when the external shape of a marker in a two-dimensional captured image is close to a parallelogram, the detected normal direction of the marker is detected as being the exact opposite direction of the actual normal direction. Therefore, the closer the shape of the marker in an image is to a parallelogram, the lower the reliability calculated compared to when the external shape is a trapezoid, making it easier to calculate the reliability appropriately. Note that parallelograms include both parallelograms with interior angles other than 90 degrees and rectangles and squares with interior angles of 90 degrees. However, if the marker is a parallelogram, the accuracy of measuring the relative position is likely to decrease not only when the external shape of the marker in a two-dimensional captured image is a parallelogram with interior angles other than 90 degrees, but also when it is a rectangle or square with interior angles of 90 degrees. Therefore, the control unit may calculate a low reliability only if the outer shape of the marker shown in the captured image is a parallelogram with an interior angle that is not 90 degrees, but may also calculate a low reliability if the outer shape of the marker shown in the captured image is a parallelogram regardless of the interior angle.

[0031] In the reliability calculation step, the control unit may calculate a lower reliability as the size of the marker in the image captured by the 2D camera becomes smaller. When the size of the marker in the image becomes smaller, the measurement accuracy of the relative position of the 2D camera with respect to the marker becomes more likely to decrease. Therefore, the smaller the size of the marker in the image, the lower the reliability calculated, making it easier to properly calculate the reliability.

[0032] This disclosure provides examples of methods for calculating the reliability of a relative position acquired based on a marker, including: (1) a method in which the closer the marker in a two-dimensional image is to the shape captured when the 2D camera is facing the marker directly, the lower the reliability calculated; (2) a method in which the closer the external shape of the marker in the image is to a parallelogram, the lower the reliability calculated; and (3) a method in which the smaller the marker in the image, the lower the reliability calculated. The control unit may calculate the reliability by combining multiple methods. For example, the control unit may weight the reliability calculated by each of the multiple methods and integrate the multiple reliability values ​​based on the weights to calculate the reliability of the acquired relative position. Furthermore, the lowest reliability value among the reliability values ​​calculated by each of the multiple methods may be used as the reliability of the relative position.

[0033] In the current position acquisition step, the control unit may acquire the current position of the traveling device without using information acquired based on markers included in the image captured by the 2D camera whose reliability does not meet a standard.

[0034] In this case, the relative position of the marker whose reliability does not satisfy the standard is not used to obtain the current position of the traveling device, so the current position of the traveling device can be obtained more accurately.

[0035] In the current position acquisition step, if there is no marker in the image captured by the 2D camera whose reliability meets the standard, the control unit may acquire the current position of the traveling device based on the driving result of the traveling drive unit.

[0036] In this case, even if the image does not include a marker whose reliability meets the standard, the current position of the traveling device can be obtained (estimated) to some extent based on the driving results of the traveling drive unit. Therefore, even if there are only a few markers placed on the path, the traveling device can easily travel automatically.

[0037] In the current position acquisition step, if multiple relative positions of the traveling device with respect to multiple markers are acquired, the control unit may acquire the current position of the traveling device based on the multiple acquired relative positions.

[0038] In this case, the current position of the traveling device can be more easily obtained with higher accuracy than when the current position of the traveling device is obtained based on the relative position of the traveling device with respect to one marker.

[0039] For example, if there are multiple markers whose reliability meets a standard in the image captured by the 2D camera, the control unit may acquire the current position of the traveling device based on the relative position of the traveling device with respect to each of the multiple markers whose reliability meets the standard. In this case, since multiple relative positions with respect to markers whose reliability meets the standard are used, the accuracy of the acquired current position is likely to be further improved.

[0040] When acquiring the current position of the traveling device based on the multiple relative positions with respect to the multiple markers, the control unit may use the reliability of each of the multiple markers. For example, the control unit may weight the relative positions with respect to the markers according to the reliability of each of the multiple markers. When the multiple current positions of the traveling device calculated based on each of the multiple relative positions differ, the control unit may acquire the final current position according to the weight of the relative positions.

[0041] However, the specific method for acquiring the current position of the traveling device can be changed. For example, when there are multiple markers in an image captured by a 2D camera, the control unit may acquire the current position of the traveling device based on the relative position of the traveling device with respect to the marker with the highest reliability among the multiple markers captured in the image. In this case, the current position of the traveling device is acquired based on the marker with the highest reliability, making it easier to acquire the current position with high accuracy.

[0042] The automated driving system and automated driving method disclosed herein may be used to automatically drive a traveling device along a path installed in a crop cultivation field. In a crop cultivation field, the environment changes due to crop growth, harvesting, soil changes, etc., so it is difficult to ensure proper automated driving using only a path map. In contrast, the automated driving system and automated driving method disclosed herein automatically drive the traveling device appropriately while minimizing the complexity of the configuration and processing by placing multiple markers at predetermined positions on the path.

[0043] <Embodiment> A typical embodiment of the present disclosure will be described below with reference to the drawings. As an example, an automatic driving system 10 (see FIG. 2) of this embodiment is used to automatically drive a traveling device 1 along a path installed in a farm. However, the automatic driving system 10 of the present disclosure can also be used to automatically drive a traveling device 1 along other paths (for example, paths installed in a factory or workshop).

[0044] (Running gear) With reference to FIG. 1, a schematic configuration of the traveling device 1 provided in the automatic traveling system 10 of this embodiment will be described. The lower left side of the page in FIG. 1 is the front of the traveling device 1. The upper right side of the page in FIG. 1 is the rear of the traveling device 1. The upper left side of the page in FIG. 1 is the right side of the traveling device 1. The lower right side of the page in FIG. 1 is the left side of the traveling device 1. The upper side of the page in FIG. 1 is the top side of the traveling device 1. The lower side of the page in FIG. 1 is the bottom side of the traveling device 1. In addition, in FIGS. 1, 3, and 6 to 8, the direction extending forward and backward from the traveling device 1 is shown as the straight traveling direction SD. When traveling straight forward or backward, the traveling device 1 travels along the straight traveling direction SD.

[0045] As shown in Fig. 1, the traveling device 1 includes a main body 2 that holds various components. In this embodiment, the main body 2 is substantially box-shaped and is formed so that the top surface is flat. An operator can stand on the top surface of the main body 2, and various objects (for example, harvested crops) can also be placed on it.

[0046] A right wheel 3R is provided on the right side of the main body 2, and a left wheel 3L is provided on the left side of the main body 2. In this embodiment, there are two right wheels 3R ​​and two left wheels 3L, but it goes without saying that the numbers of right wheels 3R ​​and left wheels 3L can be changed. Also, instead of the right wheels 3R ​​and the left wheels 3L, another configuration for running the traveling device 1 (for example, a belt-like crawler belt, etc.) may be used.

[0047] The main body 2 is provided with a travel drive unit 4 that drives the right wheel 3R and the left wheel 3L to rotate. The travel drive unit 4 can change the traveling direction of the main body 2 of the traveling device 1 and causes the main body 2 to travel in the traveling direction. As an example, the travel drive unit 4 of this embodiment can change the traveling direction of the traveling device 1 relative to the straight traveling direction SD by changing at least one of the rotation speed and rotation direction of the right wheel 3R and the left wheel 3L. Furthermore, the travel drive unit 4 can make the traveling device 1 travel forward or backward along the straight traveling direction SD by matching both the rotation speed and rotation direction of the right wheel 3R and the left wheel 3L. However, it is also possible to change the configuration for changing the traveling direction of the traveling device. For example, the travel drive unit may change the traveling direction by changing the orientation of the front wheels of the traveling device.

[0048] The main body 2 is provided with 2D cameras 5 (5F, 5B) that capture two-dimensional images (moving images in this embodiment). The 2D cameras 5 can capture images of markers 30 (see FIGS. 3 to 9) placed on the passage 100. The number of 2D cameras 5 provided on the main body 2 and the direction of the imaging optical axis will be described later.

[0049] The main body 2 is provided with a communication I / F 6 and a control unit 7. The communication I / F 6 communicatively connects the traveling device 1 with a control device 20 (see FIG. 2). The control unit 7 is responsible for drive control of the traveling drive unit 4, various controls related to the 2D camera 5, control of communication via the communication I / F 6, etc.

[0050] (Electrical configuration) With reference to FIG. 2, the electrical configuration of the autonomous driving system 10 of this embodiment will be described in outline. As an example, the autonomous driving system 10 of this embodiment includes a driving device 1 and a control device 20. The driving device 1 and the control device 20 are connected via wireless communication. In the autonomous driving system 10 of this embodiment, the CPU 22 of the control device 20 controls the autonomous driving of the driving device 1 by executing an autonomous driving control process (see FIG. 9) described below. However, it is also possible to change the method of controlling the autonomous driving of the driving device 1. For example, the CPU 22 of the control device 20 and the CPU 8 of the driving device 1 may cooperate to execute the autonomous driving control process. Alternatively, the CPU 8 of the driving device 1 may execute the autonomous driving control process independently. Alternatively, a control unit of another device may execute part of the autonomous driving control process.

[0051] As described above, the traveling device 1 includes the traveling drive unit 4, the 2D cameras 5 (5F, 5B), the communication I / F 6, and the control unit 7. The control unit 7 includes a CPU 8, which is a controller responsible for control, and a storage device 9 capable of storing programs, data, and the like. When the traveling device 1 executes at least a part of the automatic traveling control process (see FIG. 10 ), which will be described later, the storage device 9 may store at least a part of an automatic traveling control program for executing the automatic traveling control process. The communication I / F 6 communicatively connects the traveling device 1 and the control device 20.

[0052] The control device 20 includes a control unit 21, a communication I / F 24, an operation unit 25, and a display unit 26. The control unit 21 includes a CPU 22, which is a controller responsible for control, and a storage device 23 capable of storing programs, data, and the like. The storage device 23 stores at least a portion of an automatic driving control program for executing the automatic driving control process (see FIG. 10 ), which will be described later. The storage device 23 also stores map data indicating a known path 100 and the respective locations of multiple markers 30 placed on the path 100. The storage device 23 also stores a driving plan indicating a route along which the traveling device 1 will automatically travel. The communication I / F 24 communicatively connects the control device 20 and the traveling device 1. The operation unit 25 is operated by a user to input various instructions to the control device 20. The operation unit 25 can be, for example, at least one of a touch panel, various buttons, a keyboard, a mouse, and the like. The display unit 26 displays various images. Instead of the operation unit 25 and the display unit 26 provided in the control device 20, an operation unit and a display unit externally connected to the control device 20 may be used.

[0053] (Passageway where the traveling device moves automatically) An example of a path 100 along which the traveling device 1 automatically travels will be described with reference to FIG. 3. The automatic traveling system 10 of this embodiment can cause the traveling device 1 to automatically travel along a known path 100. The known path 100 illustrated in FIG. 3 includes a main path 101 and a branch path 102 branching off from the main path 101. In this embodiment, each of the multiple paths (main path 101 and branch path 102) that make up the entire path 100 is straight and has no curves. The branch path 102 intersects with the main path 101 (at a right angle in this embodiment). However, each path may be curved.

[0054] Two-dimensional markers 30 having specific geometric characteristics are placed at each of a plurality of predetermined positions on a known passage 100. Each marker 30 is placed so that its two-dimensional geometric characteristics face the inside of the passage 100. The markers 30 are photographed by a 2D camera 5 provided on the traveling device 1. When the traveling device 1 is autonomously traveling, it is necessary to accurately control the traveling direction of the traveling device 1, particularly at branch points (intersections) where multiple passages branch off. Therefore, in the example shown in FIG. 3, markers 30 are always placed near each of the multiple branch points (in this embodiment, just before and just after the traveling device 1 enters the branch point). Furthermore, even if the traveling device 1 is located along each passage, it is difficult to accurately obtain the current position of the traveling device 1 solely from the driving results of the traveling drive unit 4 (e.g., the diameters and rotation speeds of the right wheel 3R and the left wheel 3L, etc.), especially if the passage is long. Therefore, in the example shown in FIG. 3, markers 30 may also be placed along each passage.

[0055] As described above, the automated driving system 10 stores map data indicating the known path 100 and the locations of the multiple markers 30 installed on the path 100. Therefore, once the relative position of the traveling device 1 with respect to a specific marker 30 is determined, the current position of the traveling device 1 on the path 100 can be determined by applying the determined relative position to the map data.

[0056] (marker) An example of a marker 30 that can be used in the automated driving system 10 will be described with reference to FIGS. 4 and 5. In this embodiment, as shown in FIG. 4, an AR marker in which a geometric shape is drawn within a two-dimensional rectangular area (a square area in this embodiment) is used as the marker 30 to be placed on the passage 100. As a result, by processing the image captured by the 2D camera 5, it becomes easier to appropriately measure the relative position (distance and direction) of the 2D camera 5 (i.e., the traveling device 1 equipped with the 2D camera 5) with respect to the marker 30 captured in the image. In this embodiment, the geometric characteristics assigned to each of the multiple markers 30 placed on the passage 100 are unique to each marker 30. The map data indicates which marker 30 is installed at which position within the passage 100. The automated driving system 10 (in this embodiment, the CPU 22 of the control device 20) can identify which of the multiple markers 30 is the marker 30 captured in the image captured by the 2D camera 5, based on the geometric characteristics of the marker 30. The automatic driving system 10 acquires the current position of the traveling device 1 on the passage 100 by applying the relative position of the traveling device 1 with respect to the identified marker 30 on the map data.

[0057] An example of a method for acquiring the relative position of the traveling device 1 with respect to the marker 30 will be described. The automatic traveling system 10 of this embodiment (in this embodiment, the CPU 22 of the control device 20) acquires the direction of the traveling device 1 with respect to the marker 30 (i.e., the direction of the marker 30 as seen from the traveling device 1) based on the position where the marker 30 appears in the image captured by the 2D camera 5 (in this embodiment, the position where the marker 30 appears in the coordinate system of the image) and the orientation of the marker 30 appearing in the image (in this embodiment, the shape of the marker 30). In addition, the automatic traveling system 10 acquires the distance of the traveling device 1 with respect to the marker 30 (the distance between the marker 30 and the traveling device 1) based on the size of the marker 30 appearing in the image captured by the 2D camera 5.

[0058] Here, as shown in FIG. 5 , when the marker 30 is photographed from a direction oblique to the normal direction ND of the two-dimensional marker 30, the shape of the marker 30 in the image is different (trapezoidal in this embodiment) from the shape (square in this embodiment) when photographed from the normal direction ND. In this case, the normal direction ND of the marker 30 is likely to be accurately detected based on the shape of the marker 30 photographed in the image. However, through repeated trials and investigations, the inventors of the present application found that when the normal direction ND of the marker 30 is parallel to the imaging optical axis of the 2D camera 5 (for example, when the 2D camera 5 photographs the marker 30 from the front), the measurement accuracy of the relative position (particularly the direction) of the 2D camera 5 with respect to the marker 30 is likely to decrease. This is thought to be due to the phenomenon that, when the normal direction ND of the marker 30 is parallel to the imaging optical axis of the 2D camera 5, the normal direction of the marker detected based on the two-dimensional image is detected as the exact opposite direction to the actual normal direction ND.

[0059] Furthermore, the closer the external shape of the marker 30 in the two-dimensional captured image is to a parallelogram, the more likely it is that the measurement accuracy of the relative position (especially the direction) of the 2D camera 5 with respect to the marker 30 decreases. This is thought to be because, when the external shape of the marker 30 in the two-dimensional image is close to a parallelogram, a phenomenon occurs in which the normal direction of the detected marker is detected as being exactly opposite to the actual normal direction.

[0060] Based on the above findings, in the automatic traveling system 10 of this embodiment, the relationship between the installation angle of the marker 30 relative to the passage 100 and the installation angle of the 2D camera 5 relative to the traveling device 1 is adjusted.

[0061] (Installation angle of marker and 2D camera) 6 to 9, the relationship between the installation angle of the markers 30 and the installation angle of the 2D camera 5 of the traveling device 1 in the automatic traveling system 10 of this embodiment will be described. As shown in FIGS. 6 and 8, in this embodiment, when the traveling device 1 and the markers 30 are viewed from above, each of the multiple markers 30 is arranged on the passage 100 so that the normal direction ND of the two-dimensional marker 30 is within a predetermined angle range with respect to the direction in which the passage 100 extends. When the traveling device 1 is traveling straight along the passage 100 (the state shown in FIGS. 6 and 8), the angle formed by the direction of the imaging optical axis O(OF, OB) of the 2D camera 5 provided on the traveling device 1 and the normal direction of the marker 30 arranged on the passage through which the traveling device 1 is traveling straight (the main passage 101 in FIG. 6 and the branch passage 102 in FIG. 8) is within a range of 25 degrees to 65 degrees (approximately 45 degrees in this embodiment). As a result, the frequency with which the normal direction ND of the marker 30 and the imaging optical axis O (OF, OB) of the 2D camera 5 become parallel decreases. Therefore, the relative position of the 2D camera 5 (traveling device 1) with respect to the marker 30 can be measured with high accuracy based on the image captured by the 2D camera 5 while suppressing an increase in the processing load.

[0062] 6 and 8, in this embodiment, multiple markers 30 are arranged in the passageway so that the normal direction ND of the markers 30 is in the range of 80 degrees to 100 degrees (approximately 90 degrees in this embodiment) with respect to the direction in which the passageway 100 extends. Furthermore, the direction of the imaging optical axis O(OF, OB) of the 2D camera 5 provided on the traveling device 1 is tilted within the range of 25 degrees to 65 degrees (45 degrees in this embodiment) with respect to the straight direction SD extending in the front-to-rear direction from the traveling device 1. In other words, the angle θ formed by the direction of the imaging optical axis O(OF, OB) of the 2D camera 5 provided on the traveling device 1 and the straight direction SD of the traveling device 1 is adjusted to be within the range of 25 degrees to 65 degrees.

[0063] In this embodiment, the two-dimensional marker 30 can be easily installed because it is sufficient to arrange the planar direction of the marker 30 so that it is aligned with the direction in which the passage 100 extends (that is, so that the normal direction ND of the marker 30 is as perpendicular as possible to the center of the passage 100). Furthermore, unlike when the planar direction of the marker 30 is inclined with respect to the direction in which the passage 100 extends, the 2D camera 5 of the traveling device 1 can photograph the marker 30 from a position closer to the marker 30 in the traveling direction, and can also photograph the marker 30 from a position farther back than the marker 30 in the traveling direction. This makes it easier to obtain the relative position of the traveling device 1 with respect to the marker 30 with higher accuracy.

[0064] 6 to 8 are used as examples to provide a more detailed explanation. FIG. 6 is a schematic diagram of the traveling device 1 viewed from above immediately before entering the branch passage 102 from the main passage 101. In the state shown in FIG. 6, the traveling device 1 is traveling straight along the main passage 101. In this case, the angle formed between the normal direction ND of the marker 30 placed on the main passage 101 and the imaging optical axis O(OF, OB) of the 2D camera 5 of the traveling device 1 tends to fall within a range of 25 degrees to 65 degrees. Therefore, the relative position of the traveling device 1 with respect to the marker 30 can be easily measured with high accuracy. In the example shown in FIG. 6, the branch passage 102 branches off perpendicularly from the main passage 101. Therefore, in the state shown in FIG. 6, the angle formed between the normal direction ND of the marker 30 placed on the branch passage 102 and the imaging optical axis OF of the 2D camera 5F of the traveling device 1 also falls within a range of 25 degrees to 65 degrees.

[0065] FIG. 7 is a schematic diagram showing the traveling device 1 viewed from above as it begins to enter the branch passage 102 from the main passage 101. As shown in FIG. 7, while the traveling device 1 is making a curve, there is a time when the imaging optical axis of the 2D camera 5 (in FIG. 7, the imaging optical axis OB of the 2D camera 5B) becomes parallel to the normal direction ND of the marker 30. However, the time required for the traveling device 1 to turn the passage 100 is short compared to the time required for the traveling device 1 to move straight along the passage 100. Therefore, in the state shown in FIG. 7, the current position of the traveling device 1 can be determined with high accuracy by referring to the drive result of the traveling drive unit 4, without using the relative position of the traveling device 1 with respect to the marker 30.

[0066] When the traveling device 1 completes its entry into the branch passage 102 from the state shown in Fig. 7, it transitions to the state shown in Fig. 8. Even in the state shown in Fig. 8, the traveling device 1 is traveling straight through the branch passage 102, so the angle formed by the normal direction ND of the marker 30 placed in the branch passage 102 and the imaging optical axis O(OF, OB) of the 2D camera 5 of the traveling device 1 tends to fall within the range of 25 degrees to 65 degrees. Therefore, the relative position of the traveling device 1 with respect to the marker 30 can be easily measured with high accuracy.

[0067] As shown in FIGS. 6 to 8 , in this embodiment, the direction of the imaging optical axis O (OF, OB) of one or more 2D cameras 5 (two 2D cameras 5 in this embodiment) provided on the traveling apparatus 1 is tilted within a range of 25 degrees to either the left or right (only the left in this embodiment) with respect to the straight traveling direction SD extending forward and backward from the traveling apparatus 1. As a result, in this embodiment, the marker 30 on the left side of the traveling apparatus 1 is captured by the 2D camera 5. According to the automatic traveling system 10 of this embodiment, the current position of the traveling apparatus 1 is acquired based on the distance and direction (relative position) of the traveling apparatus 1 relative to the marker 30. Therefore, the current position of the traveling apparatus 1 can be appropriately acquired even if the 2D camera 5 does not capture both the markers 30 on the left and right sides of the traveling apparatus 1. Furthermore, when capturing an image of only the left or right side of the traveling apparatus 1, the number of 2D cameras 5 installed on the traveling apparatus 1 can be reduced compared to when capturing images of both the left and right sides. Therefore, automatic traveling of the traveling apparatus 1 is achieved with a simpler configuration and processing.

[0068] As shown in FIGS. 6 to 8, two 2D cameras 5F, 5B are provided on the main body 2 of the traveling apparatus 1 of this embodiment. Here, the shooting direction is either the left or right side with respect to the straight traveling direction SD extending forward and backward from the traveling apparatus 1. In this embodiment, the shooting direction is the left side of the traveling apparatus 1. The shooting optical axis OF of one of the 2D cameras 5F is tilted within a range of 25 degrees to 65 degrees toward the shooting direction (left side) with respect to the forward straight traveling direction SD. The shooting optical axis OB of the other 2D camera 5B is tilted within a range of 25 degrees to 65 degrees toward the shooting direction (left side) with respect to the rear of the straight traveling direction SD. According to this embodiment, by capturing images only on the left or right side of the traveling apparatus 1, an increase in the number of 2D cameras 5 installed on the traveling apparatus 1 and an increase in the image processing load are appropriately suppressed. Furthermore, by capturing images of the left diagonally forward and left diagonally backward (or the right diagonally forward and right diagonally backward) of the traveling device 1 by the two 2D cameras 5F, 5B, the current position of the traveling device 1 in the direction along the passage 100 can be acquired with higher accuracy. This makes it easier for the traveling device 1 to travel automatically in a more appropriate manner.

[0069] Furthermore, as described above, the closer the external shape of the marker 30 in a two-dimensional captured image is to a parallelogram, the more likely it is that the accuracy of measuring the relative position (particularly the direction) of the 2D camera 5 with respect to the marker 30 decreases. Here, if the marker 30 is captured by the 2D camera 5 from diagonally above or below the left or right of the normal direction ND of the marker 30, the external shape of the marker 30 captured in the captured image is likely to become a parallelogram. Therefore, as shown in FIG. 9 , in this embodiment, when the marker 30 is viewed from a horizontal direction perpendicular to the normal direction ND, the installation angle of the marker 30 and the installation angle of the 2D camera 5 provided on the traveling device 1 are adjusted so that the angle Φ formed between the normal direction ND of the marker 30 and the imaging optical axis O of the 2D camera 5 is 20 degrees or less, more preferably 10 degrees or less. In other words, the installation angle of the marker 30 and the installation angle of the 2D camera 5 provided on the traveling device 1 are adjusted so that the normal direction ND of the marker 30 and the imaging optical axis O of the 2D camera 5 are as parallel as possible. As a result, the relative position of the 2D camera 5 (traveling device 1) with respect to the marker 30 can be measured more accurately.

[0070] (Automatic driving control processing) The automatic driving control process executed by the automatic driving system 10 of this embodiment will be described with reference to Fig. 10. As described above, in this embodiment, the CPU 22 of the control device 20 executes the automatic driving control process. However, at least a part of the automatic driving control process may be executed by a control unit of another device (for example, the CPU 8 of the driving device 1, etc.). The CPU 22 of the control device 20 executes the automatic driving control process illustrated in Fig. 10 in accordance with an automatic driving control program stored in the storage device 23.

[0071] First, the CPU 22 acquires an image captured by the 2D camera 5 of the traveling device 1 (S1). In this embodiment, data of the image captured by the 2D camera 5 of the traveling device 1 is transmitted to the control device 20 via wireless communication. Next, the CPU 22 processes the image acquired in S1 to determine whether or not the captured image includes a marker 30 (S2). If the captured image does not include a marker 30 (S2: NO), the CPU 22 acquires (estimates) the current position of the traveling device 1 based on map data stored in the storage device 23 and the driving results of the traveling drive unit 4 (in this embodiment, the diameters and rotation speeds of the right wheel 3R and the left wheel 3L) (S11). For example, the CPU 22 may acquire the current position of the traveling device 1 by applying, to the map data, the accurate and latest position of the traveling device 1 acquired in the past based on the markers 30, etc., and a traveling route from the latest position acquired based on the driving results of the traveling drive unit 4. By performing the process of S11, even if the marker 30 does not appear in the image captured by the 2D camera 5, the current position of the traveling device 1 can be acquired (estimated) to some extent.

[0072] If the image acquired in S1 includes a marker 30 (S2: YES), the CPU 22 calculates the reliability of the relative position of the traveling device 1 acquired based on the marker 30 (the distance and direction of the traveling device 1 relative to the marker 30) based on the state of the marker 30 appearing in the image (S3).

[0073] In S3 of this embodiment, the CPU 22 calculates a lower reliability as the shape of the marker 30 in the image captured by the 2D camera 5 approaches the shape captured when the imaging optical axis O(OF, OB) of the 2D camera 5 is aligned with the normal direction ND of the marker 30 (i.e., when the 2D camera 5 is directly facing the marker 30). In this embodiment, when the 2D camera 5 is positioned directly facing the marker 30 to capture an image, the shape of the marker 30 in the image becomes square. In this case, the measurement accuracy of the relative position of the 2D camera 5 with respect to the marker 30 is likely to decrease. Therefore, the closer the shape of the marker 30 in the image approaches a square, the lower the reliability calculated, making it easier to properly calculate the reliability.

[0074] Furthermore, in S3 of this embodiment, the closer the external shape of the marker 30 shown in the image captured by the 2D camera 5 is to a parallelogram, the lower the reliability calculated by the CPU 22 compared to when the external shape is a trapezoid. As described above, the closer the external shape of the marker 30 shown in the two-dimensional captured image is to a parallelogram, the more likely it is that the measurement accuracy of the relative position (particularly the direction) of the 2D camera 5 with respect to the marker 30 will decrease. Therefore, the closer the shape of the marker 30 shown in the image is to a parallelogram, the lower the reliability calculated, making it easier to properly calculate the reliability.

[0075] Furthermore, in S3 of this embodiment, the smaller the size of the marker 30 that appears in the image captured by the 2D camera 5, the lower the reliability that the CPU 22 calculates. If the size of the marker 30 that appears in the image is small, the measurement accuracy of the relative position of the 2D camera 5 with respect to the marker 30 is likely to decrease. Therefore, the smaller the size of the marker 30 that appears in the image, the lower the reliability that is calculated, making it easier to properly calculate the reliability.

[0076] The CPU 22 may calculate the reliability by combining two or more of the three reliability calculation methods described above. For example, the CPU 22 may weight the reliability calculated by each of the multiple methods and integrate the multiple reliabilities based on the weights to calculate the reliability of the acquired relative position. Furthermore, the lowest reliability among the reliability calculated by each of the multiple methods may be adopted as the reliability of the relative position acquired based on the marker 30.

[0077] Next, the CPU 22 determines whether or not there is a marker 30 whose reliability calculated in S3 satisfies the criterion (for example, the reliability is equal to or greater than a threshold) among one or more markers 30 appearing in the image acquired in S1 (S5). If no marker 30 whose reliability satisfies the criterion is appearing in the image (S5: NO), the CPU 22 acquires the current position of the traveling device 1 based on the map data stored in the storage device 23 and the driving results of the traveling drive unit 4 (S11). Therefore, even if no marker 30 whose reliability satisfies the criterion is appearing in the image, the current position of the traveling device 1 can be acquired (estimated) to a certain extent based on the driving results of the traveling drive unit 4. This makes it easier for the traveling device 1 to travel automatically and appropriately, even when there are only a few markers 30 arranged on the path.

[0078] If a marker 30 whose reliability meets the criterion is present in the image acquired in S1 (S5: YES), the CPU 22 identifies only the markers 30 whose reliability meets the criterion from among one or more markers 30 appearing in the image acquired in S1 (S6). In the subsequent processes of S7 and S8, the current position of the traveling device 1 is acquired using only the areas above the markers 30 whose reliability meets the criterion, without using information based on markers 30 whose reliability does not meet the criterion among the markers 30 included in the image acquired in S1. As a result, the current position of the traveling device 1 is acquired with higher accuracy.

[0079] The CPU 22 acquires the relative position of the traveling device 1 (2D camera 5) with respect to the marker 30 identified in S6 (i.e., the marker 30 whose reliability meets the standard) among the markers 30 appearing in the image acquired in S1 (S7). The method of acquiring the relative position of the traveling device 1 with respect to the marker 30 is as described above. The CPU 22 acquires the current position of the traveling device 1 by matching, on the map data, the position of the marker 30 identified in S6 among the positions of the multiple markers 30 in the map data stored in the storage device 23 with the relative position with respect to the marker 30 acquired in S7 (S8). As a result, the current position of the traveling device 1 is acquired with high accuracy regardless of changes in the environment in which the passage 100 is installed, etc.

[0080] In the processes of S6 to S8, if a plurality of relative positions of the traveling apparatus 1 with respect to a plurality of markers 30 are acquired (that is, in this embodiment, if a plurality of markers 30 whose reliability meets the standard are present in the image), the CPU 22 may acquire the current position of the traveling apparatus 1 based on the acquired plurality of relative positions. For example, the CPU 22 may weight the relative positions with respect to the markers 30 according to the reliability of each of the plurality of markers 30 identified in S6. If the plurality of current positions of the traveling apparatus 1 calculated based on each of the plurality of relative positions differ, the CPU 22 may acquire the final current position according to the weight of the relative positions.

[0081] Furthermore, in the processes of S6 to S8, the CPU 22 may acquire the current position of the traveling device 1 based on the relative position of the traveling device 1 with respect to one marker 30 that has the highest reliability among the multiple markers 30 that appear in the image. In this case, the current position of the traveling device 1 is acquired based on the marker 30 with the highest reliability, which makes it easier to acquire the current position with high accuracy.

[0082] The CPU 22 controls the driving of the traveling drive unit 4 based on the current position of the traveling device 1 acquired in S8 or S11 and the traveling plan stored in the storage device 23, thereby causing the traveling device 1 to automatically travel along the passage 100 (S12). If a trigger to end the process has not been input (S14: NO), the process returns to S1, and the processes of S1 to S14 are repeated. As a result, the traveling device 1 automatically travels based on the traveling plan. When a trigger to end the process is input, the automatic traveling control process ends.

[0083] The techniques disclosed in the above embodiments are merely examples. Therefore, it is possible to modify the techniques exemplified in the above embodiments. For example, it is possible to implement only a part of the techniques exemplified in the above embodiments. For example, if the installation angle of the marker 30 and the installation angle of the 2D camera 5 provided on the traveling device 1 are adjusted so that the normal direction ND of the marker 30 and the imaging optical axis O of the 2D camera 5 are as parallel as possible, the shape of the marker 30 photographed by the 2D camera 5 is unlikely to be a parallelogram. In this case, in the process (S3) of calculating the reliability of the marker 30 depicted in the image, it is not necessary to determine whether the shape of the marker 30 depicted in the image is close to a parallelogram.

[0084] In the above embodiment, the markers 30 are arranged in the passageway so that the normal direction ND of the markers 30 is in a range of 80 degrees to 100 degrees (approximately 90 degrees in this embodiment) with respect to the direction in which the passageway 100 extends. The direction of the imaging optical axis O (OF, OB) of the 2D camera 5 provided on the traveling device 1 is tilted at an angle of 25 degrees to 65 degrees (45 degrees in this embodiment) with respect to the straight direction SD extending forward and backward from the traveling device 1. As a result, the angle between the normal direction ND of the markers 30 and the imaging optical axis O (OF, OB) of the 2D camera 5 is adjusted. However, it is also possible to change the installation angle of the markers 30 relative to the passageway 100 and the installation angle of the 2D camera 5 relative to the traveling device 1. For example, the direction of the imaging optical axis O of the 2D camera 5 provided on the traveling device 1 may be parallel to or at an angle of 90 degrees with respect to the straight direction SD extending forward and backward from the traveling device 1. The normal direction ND of the marker 30 installed on the passage 100 may be tilted within a range of 25 degrees to 65 degrees with respect to the direction in which the passage 100 extends. Even in this case, the frequency with which the normal direction ND of the marker 30 and the imaging optical axis O of the 2D camera 5 become parallel decreases. Therefore, the relative position of the 2D camera 5 (traveling device 1) with respect to the marker 30 is measured with high accuracy while suppressing an increase in the processing load. [Explanation of symbols]

[0085] 1 Running gear 2 Main unit 4 Travel drive unit 5(5F,5B) 2D camera 10 Autonomous Driving System 20 Control device 22 CPU 23 Storage device 30 markers 100 routes SD Go straight ND normal direction

Claims

1. An automatic traveling system comprising a traveling device that travels along a path and a control unit that controls the traveling of the traveling device, and automatically traveling the traveling device along the known path, a two-dimensional marker having a particular geometric characteristic is disposed at each of a plurality of predetermined positions on the passage; The traveling device is a travel drive unit that is capable of changing the traveling direction of the main body and causes the main body to travel in the traveling direction; a 2D camera provided on the main body and capable of photographing the markers arranged on the passage; Equipped with The control unit a relative position acquisition step of processing an image captured by the 2D camera to identify at least one of the markers included in the image and acquiring a distance and a direction of the traveling device relative to the identified marker; a current position acquisition step of acquiring a current position of the traveling device within the map of the passage based on map data indicating the arrangement of each of the plurality of markers on the map of the passage and the relative position of the traveling device with respect to the identified marker; a travel control step of automatically causing the traveling device to travel along the path by controlling the driving of the traveling drive unit based on the current position of the traveling device acquired in the current position acquisition step; Run When the traveling device and the plurality of markers are viewed from above, each of the plurality of markers is disposed on the passage so that a normal direction of the two-dimensional marker is within a predetermined angle range with respect to a direction in which the passage extends; An automatic driving system characterized in that, when the traveling device is traveling straight along the passage, the angle formed by the direction of the imaging optical axis of the 2D camera provided on the traveling device and the normal direction of the marker placed on the passage along which the traveling device is traveling straight is within the range of 25 degrees to 65 degrees.

2. The automatic driving system according to claim 1, A plurality of the markers are arranged in the passage so that the normal direction of the marker is within a range of 80 degrees to 100 degrees with respect to the direction in which the passage extends, An automatic driving system characterized in that the direction of the imaging optical axis of the 2D camera provided on the traveling device is tilted within a range of 25 degrees to 65 degrees with respect to the straight direction extending forward and backward from the traveling device.

3. The automatic driving system according to claim 2, An automatic driving system characterized in that the direction of the imaging optical axis of one or more of the 2D cameras provided on the traveling device is tilted within a range of 25 degrees to 65 degrees to either the left or the right with respect to the straight-line direction extending forward and backward from the traveling device.

4. The automatic driving system according to claim 3, Two of the 2D cameras are provided on the main body of the traveling device, When the photographing direction is either left or right with respect to the straight direction extending forward and backward from the traveling device, the direction of the photographing optical axis of one of the 2D cameras is tilted within a range of 25 degrees to 65 degrees toward the photographing direction with respect to the forward straight-ahead direction, An automatic driving system characterized in that the direction of the imaging optical axis of the other 2D camera is tilted within a range of 25 degrees to 65 degrees toward the imaging direction relative to the rear of the straight-ahead direction.

5. The automatic driving system according to claim 1, The control unit An automated driving system further comprising a reliability calculation step of calculating the reliability of the relative position obtained based on the marker based on the state of the marker captured in the image taken by the 2D camera.

6. The automatic driving system according to claim 5, An autonomous driving system characterized in that, in the reliability calculation step, the control unit calculates a lower reliability the closer the shape of the marker in the image captured by the 2D camera is to the shape captured when the imaging optical axis of the 2D camera is aligned with the normal direction of the marker.

7. The automatic driving system according to claim 5, The marker is two-dimensional, and the outer shape of a region in which a geometric feature is formed is rectangular, The control unit, in the reliability calculation step, calculates a lower reliability when the outer shape of the marker in the image captured by the 2D camera is closer to a parallelogram than when the outer shape is a trapezoid.

8. The automatic driving system according to claim 5, In the current position acquisition step, the control unit An automated driving system characterized by obtaining the current position of the driving device without using information obtained based on markers included in images captured by the 2D camera whose reliability does not meet the standard.

9. The automatic driving system according to claim 8, In the current position acquisition step, the control unit An automatic driving system characterized in that, when the image captured by the 2D camera does not contain a marker whose reliability meets the standard, the current position of the driving device is obtained based on the driving results of the driving drive unit.

10. The automatic driving system according to claim 1, In the current position acquisition step, the control unit An automatic driving system characterized in that, when multiple relative positions of the traveling device are obtained with respect to multiple markers, the current position of the traveling device is obtained based on the multiple obtained relative positions.

11. An automatic traveling method executed by an automatic traveling system that includes a traveling device that travels along a path and a control unit that controls traveling of the traveling device, and that automatically travels the traveling device along the known path, a two-dimensional marker having a particular geometric characteristic is disposed at each of a plurality of predetermined positions on the passage; The traveling device is a travel drive unit that is capable of changing the traveling direction of the main body and causes the main body to travel in the traveling direction; a 2D camera provided on the main body and capable of photographing the markers arranged on the passage; Equipped with The automatic driving method includes: a relative position acquisition step of processing an image captured by the 2D camera to identify at least one of the markers included in the image and acquiring a distance and a direction of the traveling device relative to the identified marker; a current position acquisition step of acquiring a current position of the traveling device within the map of the passage based on map data indicating the arrangement of each of the plurality of markers on the map of the passage and the relative position of the traveling device with respect to the identified marker; a travel control step of automatically causing the traveling device to travel along the path by controlling the driving of the traveling drive unit based on the current position of the traveling device acquired in the current position acquisition step; Including, When the traveling device and the plurality of markers are viewed from above, each of the plurality of markers is disposed on the passage so that a normal direction of the two-dimensional marker is within a predetermined angle range with respect to a direction in which the passage extends; An automatic driving method characterized in that, when the traveling device is traveling straight along the passage, the angle formed by the direction of the imaging optical axis of the 2D camera provided on the traveling device and the normal direction of the marker placed on the passage along which the traveling device is traveling straight is within a range of 25 degrees to 65 degrees.

Images