Method for Drawing Map, Map Drawing Device, Computer Program, Computer Readable Medium, and Vehicle

Abstract

When an autonomous vehicle plans a new driving mission in a restricted area, a first set of environmental information is obtained by a device flying above the restricted area including the vehicle's starting point and destination point, and a second set of environmental information is obtained after a time interval, each set of environmental information including the positions and geometries of objects in the restricted area. Then, objects present at the same positions in the first set of environmental information and the second set of environmental information are classified as semi-static objects, and objects present at different positions in the first set of environmental information and the second set of environmental information are classified as dynamic objects. A map including the positions and geometries of each object classified as a semi-static object is then created.

Description

Technical Field

[0001] The present invention relates to a method, an apparatus, a computer program, and a computer-readable medium for mapping an area in which an autonomous vehicle drives, and also relates to a vehicle. Background Art

[0002] When an autonomous vehicle plans a new driving task, first, a route from a starting point to a destination point is set, and then, the autonomous vehicle travels along the route. To set the route, it is necessary to be able to recognize at least a map of an available area for driving the vehicle, and it is desirable that such a map has higher accuracy. As an example of a technique for improving the accuracy of the map, a technique has been proposed in which, based on observation information and position information obtained from an image around the vehicle captured by a camera, attribute information (attribute map) indicating the attributes of each of a plurality of areas (obtained by dividing around the vehicle) is created, and thus the reliability of the attributes in each area is calculated from a plurality of sets of attribute information corresponding to the same position. According to this technique, the reliability of the attributes of an area is set to be larger because the number of information sets indicating the same attribute among a plurality of sets of attribute information included in the same area is larger.

[0003] Cited Literature

[0004] PTL1: JP2017-215940A Summary of the Invention

[0005] Technical Problem

[0006] In recent years, autonomous vehicles have been used for industrial purposes. In restricted areas for industrial use, such as container loading and unloading sites at docks in ports, construction sites, and factory sites, the area available for driving a vehicle changes over time because, for example, the positions of containers or sand piles in the area change over time. However, in the above-described conventional technique, the reliability of attributes is calculated based on the assumption that objects having the same attributes generally always exist in the same position. Therefore, it is difficult to create an appropriate map in such an environment where the available area for driving changes over time, such as in a restricted area for industrial use.

[0007] Therefore, an object of the present invention is to provide a method for mapping, a mapping device, a computer program, a computer-readable medium, and a vehicle that can specify an area available for driving an autonomous vehicle even in a restricted area for industrial use.

[0008] Technical Solution

[0009] According to the present invention, a map of a restricted area in which an autonomous vehicle drives is created. That is, when the autonomous vehicle plans a new driving task, a first set of environmental information is obtained by a device flying above the restricted area including the driving starting point and the driving destination point of the autonomous vehicle, and a second set of environmental information is obtained after a time interval. Each set of environmental information includes the position and geometry of objects in the restricted area. Then, the objects existing at the same position in the first set of environmental information and the second set of environmental information are classified as semi-static objects, and the objects existing at different positions in the first set of environmental information and the second set of environmental information are classified as dynamic objects. Then, a map including the position and geometry of each object classified as a semi-static object is created.

[0010] Function of the present invention

[0011] According to the present invention, even in a restricted area for industrial use, an area available for driving an autonomous vehicle can be specified. Description of the drawings

[0012] Figure 1 It is an illustrative view showing an example of the state of the restricted area viewed from above.

[0013] Figure 2 It is an illustrative view showing an example of another state of the restricted area viewed from above.

[0014] Figure 3 It is a block diagram showing an example of the overall configuration of the system.

[0015] Figure 4 It is a flowchart showing an example of the process executed in the entire system.

[0016] Figure 5 It is a flowchart showing an example of the map drawing process.

[0017] Figure 6 It is a flowchart showing an example of the preprocessing process.

[0018] Figure 7 It is a flowchart showing an example of the first analysis process.

[0019] Figure 8 It is a flowchart showing an example of the second analysis process.

[0020] Figure 9 It is a flowchart showing an example of the classification and map creation process.

[0021] Figure 10 It is an explanatory diagram showing an example of a set of environmental information of the restricted area.

[0022] Figure 11An explanatory diagram showing an example of another set of environmental information of a restricted area.

[0023] Figure 12 An explanatory diagram showing an example of a static map.

[0024] Figure 13 An explanatory diagram showing an example of a dynamic map.

[0025] Figure 14 An explanatory diagram showing another example of a dynamic map.

[0026] Figure 15 An explanatory diagram showing another example of a dynamic map.

[0027] Figure 16 An explanatory diagram showing an example of the configuration of the hardware of a computer. Detailed implementation

[0028] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0029] In an embodiment, a system will be described that is capable of creating a map for setting a driving route of an autonomous vehicle in a restricted area for industrial use and enabling the autonomous vehicle to travel along a route created based on the map. In this embodiment, an example where the restricted area for industrial use is a port area for loading and unloading containers and the like will be described. As used herein, the term "autonomous vehicle" refers to a vehicle that performs so-called fully autonomous driving control capable of driving without a driver, as well as a vehicle that performs autonomous driving control or driving assistance control during manned driving.

[0030] First, an overview of this embodiment will be described. Figure 1 An illustrative view showing the position and geometry of an object in a restricted area at a certain time as viewed from above. In Figure 1 , each square colored in solid color or black indicates a building in the restricted area, and each shaded square indicates goods such as containers. Here, Figure 1 , moving bodies such as vehicles are omitted. In Figure 1 In the state shown, when the autonomous vehicle moves from point A to point B, the autonomous vehicle can travel along the driving route indicated by the dotted line, for example, as long as there are no other vehicles obstructing the travel.

[0031] On the other hand, Figure 2 is an illustrative view from above showing the position and geometry of an object in a restricted area at a time different from the time in Figure 1 . Comparing Figure 1 and Figure 2, move some containers, etc., so that the area in the restricted area available for driving a vehicle changes. Thus, when moving from point A to point B in the restricted area in Figure 2 , the autonomous vehicle cannot drive along the same driving route as that found at the timing in Figure 1 .

[0032] Here, the driving route generated when the autonomous vehicle plans a new driving task is based on a map of an area including the driving starting point and the driving destination point of the autonomous vehicle. Generally, for public roads, a driving route can be generated based on the same map at any time because the available roads for vehicles do not change. However, as Figure 1 and Figure 2 show, in such an environment where the available area of the vehicle changes due to the movement of containers, etc., such as a restricted area for industrial use, an appropriate driving route cannot be generated by a method similar to that for public roads. A map-drawing technique that can generate an appropriate driving route even in such an environment will be described according to the present embodiment. Hereinafter, the map obtained by this map-drawing technique is referred to as a "dynamic map".

[0033] Figure 3 is a block diagram showing the overall configuration of a system 100 according to the present embodiment. The system 100 includes a truck 10, a server 20, and a drone 30.

[0034] The truck 10 is an example of an autonomous vehicle and travels in a restricted area to transport containers, etc. In the present invention, the autonomous vehicle is not limited to a truck, but includes all types of vehicles that travel in a restricted area, such as industrial vehicles including forklifts and tractors, and engineering vehicles.

[0035] The server 20 is a computer capable of communicating with the truck 10 and the drone 30, and creates a dynamic map indicating the available area for driving in the restricted area at that time in response to a request from the truck 10.

[0036] The drone 30 is an example of a device that flies above the restricted area and obtains environment information including information about objects in the restricted area. Information about an object includes the position and geometry (e.g., shape and size) of the object.

[0037] The truck 10 includes an autonomous driving unit 11 that executes various controls related to the autonomous driving function. For example, the autonomous driving unit 11 is implemented in a preset electronic control unit (ECU) installed in the truck 10. The autonomous driving unit 11 includes a driving planning unit 12, a route generation unit 13, an autonomous driving control unit 14, a communication unit 15, and a dynamic map storage unit 16.

[0038] The driving planning unit 12 receives a driving instruction specifying a starting point and a destination from an external system (not shown), and plans a new driving task based on the driving instruction. Then, the driving planning unit 12 provides the driving task to the route generation unit 13.

[0039] The route generation unit 13 requests a dynamic map indicating an available area for driving in a restricted area at that time from the server 20, and receives the dynamic map from the server 20. Then, the route generation unit 13 generates a driving route for the truck 10 from the starting point to the destination based on the dynamic map.

[0040] The autonomous driving control unit 14 performs autonomous driving control so that the truck 10 travels along the driving route generated by the route generation unit 13. Specifically, the autonomous driving control unit 14 determines the driving direction of the truck 10 while receiving the position of the truck 10 through GPS (Global Positioning System). Then, the autonomous driving control unit 14 outputs a control signal to the engine ECU to generate wheel drive torque, and outputs a control signal to the steering ECU to perform steering control. In addition, for example, when the vehicle reaches the destination, or when various sensing devices installed in the truck 10 detect an obstacle, the autonomous driving control unit 14 outputs a control signal to the brake ECU as needed.

[0041] The communication unit 15 is a device for communicating with the truck 10 and the outside, and performs modulation and demodulation of radio waves transmitted and received via an antenna. Wireless communication can be performed by a freely selectable method (such as a standardized method including various wireless local area networks (LANs) specified in IEEE (Institute of Electrical and Electronics Engineers) 802.11 and other non-standardized methods). The truck 10 performs wireless communication with at least the external system providing the driving instruction and the server 20.

[0042] The dynamic map storage unit 16 is a memory that stores the dynamic map received from the server 20.

[0043] The server 20 includes an environment information acquisition unit 21, a data analysis unit 22, a classification unit 23, a map creation unit 24, a communication unit 25, an environment information storage unit 26, an analysis standard storage unit 27, a static map storage unit 28, and a dynamic map storage unit 29.

[0044] The environment information acquisition unit 21 receives a request for a driving route from the truck 10, and sends a request for environment information of the restricted area to the drone 30. Then, the environment information acquisition unit 21 receives the environment information from the drone 30, and stores the environment information in the environment information storage unit 26.

[0045] The data analysis unit 22 performs a preprocessing process and an analysis process on the environmental information received from the drone 30 to detect an object in a restricted area and obtain information about the object.

[0046] The classification unit 23 classifies each detected object. Specifically, an object that does not move is classified as a "static object", an object that can move after a period of time is classified as a "semi-static object", and an object that can move continuously is classified as a "dynamic object". In a port area, examples of static objects can include buildings such as administrative offices, examples of semi-static objects can include goods such as containers, and examples of dynamic objects can include transportation vehicles such as trucks.

[0047] The map creation unit 24 creates a dynamic map including the positions and geometries of at least semi-static objects based on the object classification performed by the classification unit 23.

[0048] Similar to the communication unit 15 of the truck 10, the communication unit 25 is a device for communicating with the server 20 and the outside, and performs modulation and demodulation of radio waves transmitted and received via an antenna. As described above, wireless communication can be performed by a freely selectable method. The server 20 performs wireless communication with at least the truck 10 and the drone 30.

[0049] The environmental information storage unit 26 is a memory that stores the environmental information received from the drone 30.

[0050] The analysis criterion storage unit 27 is a memory in which various types of information used as criteria in the analysis of the environmental information performed in the data analysis unit 22 are stored.

[0051] The static map storage unit 28 is a memory that pre-stores a static map. The static map records the positions and geometries of each static object in the restricted area.

[0052] The dynamic map storage unit 29 is a memory that stores the dynamic map created by the map creation unit 24.

[0053] The drone 30 includes sensing devices such as a camera 31, a radar 32, and a lidar 33, and also includes a communication unit 34.

[0054] For example, the camera 31 can be a digital camera or an infrared camera using an image sensor such as a CCD or a CMOS. The camera 31 can also be a stereo camera configured to obtain 3D stereo vision, or a TOF (time of flight) camera also known as a flash lidar. The camera 31 is installed at a position capable of capturing an image of the view downward from the drone 30, and captures an image of an area including the restricted area from above.

[0055] The radar 32 is a device that uses electromagnetic waves (such as millimeter waves). The radar 32 is installed at a position capable of emitting electromagnetic waves downward from the drone 30. The radar 32 emits electromagnetic waves and senses the reflected waves of the emitted electromagnetic waves reflected by an object on the ground to detect the position, geometry, etc. of the object.

[0056] The lidar 33 is a device that uses pulsed waves of laser, which is a short-wavelength electromagnetic radiation such as ultraviolet light, visible light, and near-infrared light. The lidar 33 is installed at a position capable of emitting electromagnetic waves downward from the drone 30. The lidar 33 emits electromagnetic waves and senses the reflected waves of the emitted electromagnetic waves reflected by an object to detect the position, geometry, etc. of the object with high precision.

[0057] Specifically, in the present embodiment, the environmental information is at least one of the image data captured by the camera 31 and the point cloud data generated based on the measurement of three-dimensional points based on the reflected waves sensed by at least one of the radar 32 and the lidar 33.

[0058] Similar to the communication unit 15 of the truck 10, the communication unit 34 is a device for communicating with the drone 30 and the outside, and modulates and demodulates the radio waves transmitted and received via the antenna. As described above, wireless communication can be performed by a freely selectable method. The drone 30 performs wireless communication with at least the server 20.

[0059] Figure 4 is a flowchart showing the processes executed in the truck 10, the server 20, and the drone 30 of the system 100.

[0060] In step 1 (abbreviated as "S1" in the figure, which also applies hereinafter), the driving planning unit 12 of the truck 10 receives a driving instruction specifying a starting point and a destination from an external system. Then, the driving planning unit 12 plans a new driving task based on the driving instruction and provides the driving task to the route generation unit 13.

[0061] In step 2, the route generation unit 13 of the truck 10 refers to the dynamic map storage unit 16 and determines whether the most recent dynamic map of the restricted area has been created within the reference period based on the creation date and time of the most recent dynamic map. Herein, the reference period is determined based on the movement frequency of semi-static objects in the restricted area. For example, when the restricted area is a port area for loading and unloading containers, etc., as in this embodiment, the semi-static object can generally be a container. Thus, for example, when the average time interval of container movement in the restricted area is 30 minutes, the reference period can be set to 30 minutes, and the route generation unit 13 can determine whether the most recent dynamic map has been created within the last 30 minutes. If it is determined that the most recent dynamic map has been created within the reference period (Yes), the process proceeds to step 9; otherwise (No), the process proceeds to step 3.

[0062] In step 3, the drone 30 takes off from the standby base and flies to a point in the sky where it can view the area including the restricted area. The drone 30 can fly autonomously to this point or can fly by manual operation. More specifically, although not shown in the Figure 4 flowchart, the truck 10 requests the server 20 to create a dynamic map, and the environment information acquisition unit 21 of the server 20 sends a signal requesting the drone 30 to acquire environment information to the drone 30. In response to this signal, the drone 30 takes off from the standby base. Alternatively, the truck 10 can directly send a signal requesting the drone 30 to acquire environment information to the drone 30.

[0063] In step 4, the drone 30 acquires the environment information of the area including the restricted area by using at least one of the camera 31, the radar 32, and the lidar 33, and each of the camera 31, the radar 32, and the lidar 33 is a sensing device. Here, the drone 30 acquires two sets of environment information at time intervals. Specifically, the drone 30 acquires the first set of environment information and acquires the second set of environment information after the time interval. In this embodiment, it is assumed that the time interval is about several seconds, but this is not limited thereto. In addition, the number of times of acquiring environment information is not limited to two and can be any number as long as it is a multiple. The drone 30 sends the two sets of acquired environment information to the server 20. Then, the environment information acquisition unit 21 of the server 20 receives the two sets of environment information from the drone 30.

[0064] In step 5, the data analysis unit 22, the classification unit 23, and the map creation unit 24 of the server 20 execute a mapping process for creating a dynamic map. The details of the mapping process will be described later.

[0065] In step 6, the map creation unit 24 of the server 20 determines whether a valid dynamic map has been created. If it is determined that a valid dynamic map has been created (Yes), the process proceeds to step 7; otherwise (No), the process returns to step 4.

[0066] In step 7, the map creation unit 24 of the server 20 stores the created dynamic map in the dynamic map storage unit 29 and also sends the created dynamic map to the truck 10. At this time, information indicating the creation date and time of the dynamic map is added to the dynamic map. In addition, the map creation unit 24 of the server 20 sends a command to return to the standby base to the drone 30.

[0067] In step 8, the drone 30 lands at the standby base.

[0068] In step 9, the route generation unit 13 of the truck 10 stores the dynamic map received from the server 20 together with the information indicating the creation date and time of the dynamic map in the dynamic map storage unit 16. Then, the route generation unit 13 generates the optimal driving route from the starting point to the destination based on the dynamic map.

[0069] In step 10, the autonomous driving control unit 14 of the truck 10 performs autonomous driving control so that the truck 10 travels along the driving route generated by the route generation unit 13.

[0070] Figure 5 It is a flowchart of the map drawing process performed by the data analysis unit 22, classification unit 23, and map creation unit 24 of the server 20.

[0071] Steps 11 to 13 are performed for each set of environmental information received from the drone 30.

[0072] In step 11, the data analysis unit 22 performs a preprocessing process on the environmental information to facilitate analysis.

[0073] In step 12, the data analysis unit 22 performs a first analysis process for detecting objects that may affect the driving of the truck 10 based on the height information of each object included in the environmental information.

[0074] In step 13, the data analysis unit 22 performs a second analysis process for detecting objects that may affect the driving of the truck 10 based on the geometric shape information of each object included in the environmental information.

[0075] In step 14, the classification unit 23 and the map creation unit 24 perform a classification and map creation process based on the data analysis result of the environmental information. The classification and map creation process classifies each object in the restricted area into a static object, a semi-static object, or a dynamic object and creates a map of the environmental information.

[0076] Figure 6 It is a flowchart of a preprocessing process executed by the data analysis unit 22 of the server 20.

[0077] In step 21, the data analysis unit 22 optionally removes noise, background, etc. from the environmental information received from the drone 30.

[0078] In step 22, the data analysis unit 22 optionally performs a smoothing process on the environmental information.

[0079] In step 23, the data analysis unit 22 determines the edge of the restricted area from the entire area indicated in the environmental information. For example, for an image captured by the camera 31, the edge of the restricted area can be determined by pre-storing the pattern of the object existing outside the edge of the restricted area and then matching the object pattern with the objects in the entire image captured by the camera 31. For example, for the point cloud data obtained by the radar 32 or the lidar 33, the edge of the restricted area can be determined by pre-storing the information of the absolute coordinates (absolute positions) of the restricted area in the analysis standard storage unit 27 and matching the coordinates of each point of the point cloud data with the absolute coordinates of the restricted area. In the following process, only the environmental information inside the edge of the restricted area is processed.

[0080] In step 24, the data analysis unit 22 stores the environmental information processed in steps 21 to 23 in the environmental information storage unit 26.

[0081] Figure 7 It is a flowchart of the first analysis process executed by the data analysis unit 22 of the server 20. The first analysis process is usually performed on the point cloud data obtained by the radar 32 and the lidar 33.

[0082] The following steps 31 to 33 are performed for each point of the point cloud data.

[0083] In step 31, the data analysis unit 22 determines whether the height position of the point to be processed is greater than or equal to a preset value based on the coordinate information of the point. If it is determined that the height position of the point is greater than or equal to the preset value (yes), the process proceeds to step 32; otherwise (no), the process proceeds to step 33.

[0084] In step 32, the data analysis unit 22 determines that there is an object with a height that may obstruct the driving of the truck 10 at the position of the point and records it in the environmental information storage unit 26.

[0085] On the contrary, in step 33, the data analysis unit 22 determines that there is no object with a height that may obstruct the driving of the truck 10 at the position of the point and records it in the environmental information storage unit 26.

[0086] In this way, the first analysis process determines the points with heights that may obstruct vehicle driving in the restricted area based on the point cloud data obtained by the radar 32 or the lidar 33. Therefore, the obstruction objects located in the restricted area can be detected.

[0087] Figure 8 is a flowchart of the second analysis process executed by the data analysis unit 22 of the server 20. The second analysis process can be performed on both the image data obtained by the camera 31 and the point cloud data obtained by the radar 32 or the lidar 33.

[0088] In step 41, the data analysis unit 22 performs segmentation on the environmental information to detect the objects indicated in the environmental information. For example, for the image data, segmentation can be performed by using methods such as a brightness threshold method or an edge determination method. For the point cloud data, a process corresponding to segmentation can be performed by determining the points with heights greater than or equal to a preset value.

[0089] The following steps 42 and 43 are performed for each object detected by the segmentation.

[0090] In step 42, the data analysis unit 22 extracts the features of the object from the environmental information, such as shape, size, and color.

[0091] In step 43, the data analysis unit 22 identifies the type of the object based on the extracted features and records the identified type in the environmental information storage unit 26. For example, the type of the object can be identified by referring to the information associated with the features and types of the object and pre-recorded in the analysis standard storage unit 27, and by matching the recorded features of the object with the extracted features. At this time, for the image data, the matching of the features can be performed based on two-dimensional information. For example, this enables the identification of whether the object shown in the image is a container, a building, or a sign drawn on the ground. For the point cloud data, the matching of the features can be based on three-dimensional information.

[0092] In this way, the second analysis process detects objects in the restricted area based on the environmental information, and then determines the area where the objects that may obstruct the vehicle driving are located in the restricted area by identifying the type of each object, even from the image data. For example, in the case where the object shown in the image is identified as a container, the points in the restricted area where the object is located are determined as the points where the vehicle cannot drive. In the case where the object shown in the image is identified as a sign drawn on the ground, the points in the restricted area where the object is located are determined as the points where the vehicle can drive. In addition, for the point cloud data, although the points in the restricted area where the objects that may obstruct the vehicle driving are located can be determined by the first analysis process or by the second analysis process, the type of each object can be further identified by the second analysis process.

[0093] Figure 9 It is a flowchart of the classification and map creation process executed by the classification unit 23 and the map creation unit 24 of the server 20.

[0094] In step 51, the classification unit 23 reads the static map stored in the static map storage unit 28.

[0095] The following steps 52 to 56 are performed for each object indicated in the environmental information. In steps 52 to 56, one of the two sets of environmental information is processed as the target.

[0096] In step 52, the classification unit 23 determines whether the object in the target environmental information is recorded in the static map. Specifically, the classification unit 23 determines whether an object having the same geometric shape at the same position as the object indicated in the target environmental information group is recorded in the static map. If it is determined that such an object is recorded in the static map (Yes), the process proceeds to step 53; otherwise (No), the process proceeds to step 54.

[0097] In step 53, the classification unit 23 classifies the object as a static object.

[0098] In step 54, the classification unit 23 determines whether the object has moved by comparing two sets of environmental information. Specifically, the classification unit 23 compares two sets of environmental information and determines whether an object with the same geometric shape at the same position as the object in the target environmental information set is indicated in another set of environmental information that is not the processing target. Then, if both sets of environmental information indicate an object with the same geometric shape at the same position, the classification unit 23 determines that the object has not moved; otherwise, it determines that the object has moved. If it is determined that the object has not moved (No), the process proceeds to step 55; otherwise (Yes), the process proceeds to step 56. Although in this embodiment, it is assumed that the number of sets of environmental information is two, when the number of sets is three or more, the classification unit 23 compares all sets of environmental information and determines whether an object with the same geometric shape at the same position as the object in the target environmental information set is indicated in each set of environmental information that is not the processing target. Then, if it is determined that all sets indicate an object with the same geometric shape at the same position, the process proceeds to step 55; otherwise, the process proceeds to step 56.

[0099] In step 55, the classification unit 23 classifies the object as a semi-static object.

[0100] In step 56, the classification unit 23 classifies the object as a dynamic object.

[0101] In step 57, based on the classification results of each object classified in steps 52 to 56, the map creation unit 24 creates a dynamic map including the positions and geometries of the objects classified as static objects and the positions and geometries of the objects classified as semi-static objects. Specifically, the map creation unit 24 creates a dynamic map of the restricted area based on the positions and geometries of each object included in the pre-processed environmental information stored in the environmental information storage unit 26. Here, although the environmental information includes information on static, semi-static, and dynamic objects, the dynamic map of this embodiment only indicates static and semi-static objects because when the truck 10 travels along the driving route after receiving the dynamic map, the dynamic objects may have moved. It should be noted that the truck 10 is equipped with an on-vehicle obstacle detection system for avoiding collisions between the truck 10 and dynamic objects.

[0102] Next, the classification processes of static, semi-static, and dynamic objects in the above steps 52 to 56 will be described with specific examples.

[0103] Figure 10 and Figure 11It is an explanatory view showing an example of two sets of acquired environmental information of a restricted area including point A and point B obtained by the unmanned aerial vehicle 30 when planning a driving task. Here, point A where the truck 10 is located is the designated starting point of the driving task, and point B is the designated destination point of the driving task. Figure 10 shows the first set of environmental information, and Figure 11 shows the second set of environmental information obtained after a time interval. Here, Figure 10 and Figure 11 The two sets of environmental information shown in have undergone a preprocessing process, a first analysis process, and a second analysis process, so that all objects in the restricted area have been detected, and the features (position and geometry) of each object have been extracted. Figure 12 shows an example of a static map.

[0104] For Figure 10 each object indicated in the first set of environmental information in, the classification unit 23 first refers to Figure 12 the static map shown in, and determines whether an object having the same geometry at the same position is recorded in the static map. Then, for Figure 10 each of the objects 40A, 40B, and 40C shown in, it is determined that an object having the same geometry at the same position is recorded in the static map, and thus, the objects 40A, 40B, and 40C are classified as static objects.

[0105] In addition, for Figure 10 the other objects in the first environmental information that are not classified as static objects, the classification unit 23 compares the objects indicated in Figure 11 the second set of environmental information in, and determines whether an object having the same geometry at the same position is indicated in the second set of environmental information. Then, for Figure 10 each of the objects 50A, 50B, 50C, and 50D shown in the first set of environmental information in Figure 11 it is determined that no object having the same geometry at the same position is indicated in Figure 11 the second set of environmental information in, and thus, the objects 50A, 50B, 50C, and 50D are classified as dynamic objects. Then, the classification unit 23 classifies the remaining objects as semi-static objects.

[0106] Figure 13This is an example of a dynamic map created by the map creation unit 24 based on the classification result. Here, the dynamic map does not indicate dynamic objects, i.e., objects 50A, 50B, 50C, and 50D. Objects 40A, 40B, and 40C classified as static objects are colored with solid colors, or black, and other objects classified as semi-static objects are represented by shading. The remaining area where the vehicle can travel is shown in white. Since this dynamic map is created in the server 20 and then sent to the truck 10, the route generation unit 13 of the truck 10 can generate a driving route along which the truck 10 can move from point A to point B while avoiding static objects and semi-static objects.

[0107] Here, the truck 10 is also classified as a semi-static object because it is in a stopped state when two sets of environmental information are obtained. However, since the truck 10 obtains its own position information, the route generation unit 13 of the truck 10 can recognize that the semi-static object at its own position in the map is the truck 10.

[0108] According to the present embodiment, when the truck 10 plans a new driving task (the truck 10 wants to start driving), all objects in the restricted area for industrial use are classified as static objects, semi-static objects, or dynamic objects, and a dynamic map including the positions and geometries of the static objects and semi-static objects is created. When the truck 10 plans a new driving task, by using the dynamic map, not only the positions of the static objects can be specified, but also the positions of the semi-static objects can be specified, so that the available area for driving the truck 10 can be specified, even for a restricted area where the available area for driving changes over time. Therefore, by creating such a dynamic map, an appropriate driving route can be generated for an autonomous vehicle in a restricted area.

[0109] In addition, according to the present embodiment, the drone 30 flies into the sky and obtains multiple sets of environmental information, and the server 20 creates a dynamic map only when the most recent dynamic map has not been created within a reference period determined based on the movement frequency of the semi-static objects. Therefore, unnecessary maps can be prevented from being created, thereby reducing the processing load of the entire system.

[0110] In addition, there may be a situation where semi-static objects in the restricted area have different movement frequencies depending on the type of semi-static object (e.g., a container or other large cargo). Therefore, the reference period can be determined based on the highest movement frequency (i.e., the shortest movement distance) obtained for each type of semi-static object in the restricted area.

[0111] Although Figure 13shows an example of a dynamic map created by the map creation unit 24 based on the classification result of the classification unit 23. However, the map creation unit 23 can further include in the dynamic map the position and geometry of each object classified as a dynamic object, as Figure 14 shown, which shows another example of a dynamic map. At this time, the dynamic map can indicate the dynamic objects indicated in either one of the two sets of environmental information, or can indicate the dynamic objects indicated in the two sets of environmental information, as Figure 14 shown. Although different from static objects and semi-static objects, the point where there is a dynamic object may be available for driving when the truck 10 drives through that point, but the probability that another dynamic object (such as another transport vehicle traveling along the same driving route) will be at the same point is relatively high. Therefore, by indicating in the dynamic map the points where there are dynamic objects, a driving route that avoids that point as much as possible can be generated.

[0112] In addition, when creating such a dynamic map that also indicates dynamic objects, as Figure 15 shown, only the dynamic objects with a small moving distance when comparing the two sets of environmental information can be indicated in the dynamic map, such as Figure 10 the object 50D in. Specifically, for the objects indicated in the first set of environmental information in Figure 10 , when there is no object with the same geometry at the same position indicated in the second set of environmental information in Figure 11 , the classification unit 23 can determine whether there is an object with the same geometry within a preset distance from the object. In other words, the classification unit 23 determines whether the moving distance of the objects obtained by comparing the first set of environmental information in Figure 10 and the second set of environmental information in Figure 11 is less than or equal to a preset value. Then, the map creation unit 24 can further indicate such dynamic objects that meet this condition in the dynamic map. Therefore, when generating the driving route of the truck 10, a driving route that avoids slow-moving dynamic objects as much as possible can be generated because such dynamic objects may obstruct the driving of the truck 10.

[0113] To make static objects, semi-static objects, and dynamic objects distinguishable in the dynamic map, static objects, semi-static objects, and dynamic objects can be depicted in different colors or patterns. For example, as Figures 13 to 15 shown. Alternatively, as another example, a dynamic map can be created using layers, and static objects, semi-static objects, and dynamic objects can be depicted on different layers.

[0114] In addition, although in this embodiment, classification of static objects is performed by referring to a static map, the present invention is not limited thereto, and classification of static objects may be omitted. That is, when comparing two sets of environmental information, both static objects and semi-static objects can be regarded as objects that have not moved, and a dynamic map can be created by only distinguishing these objects from dynamic objects.

[0115] In addition, when classifying objects between semi-static objects and dynamic objects, the classification unit 23 can specify the type of the object (e.g., a container or a transport vehicle). To specify the type, the processing result of step 43 described above can be used. Then, when classifying objects between semi-static objects or dynamic objects, the classification unit 23 refers to the information that is pre-stored in the analysis criterion storage unit 27 and indicates whether the type of each object corresponds to a semi-static object or a dynamic object, and the classification unit 23 determines whether the type of the object (i.e., the object that should be classified as a semi-static object) existing at the same position in multiple sets of environmental information corresponds to a dynamic object. Then, when the object meets this condition, the map creation unit 24 can create a dynamic map with a notification indicating the possibility that the object will move added thereto. For example, when comparing Figure 10 and 11 when, Figure 10 the object 60 indicated in the first set of environmental information in

[0116] Figure 16 does not move, and thus, according to the above process, the object 60 is classified as a semi-static object. However, the type of the object 60 is actually a transport vehicle, and the object 60 should generally be classified as a dynamic object. In this case, the map creation unit 24 can add a notification indicating this to the dynamic map.

[0117] FIG. is an explanatory diagram showing an example of the configuration of the hardware of the computer in the server 20. The computer includes a processor 201, a RAM 202, a ROM 203, a portable storage medium drive 204, an input / output device 205, and a communication interface 206.

[0118] The RAM 202 includes, for example, DRAM and SRAM, which are volatile memories where data is lost when the power supply is interrupted. Programs are loaded into the RAM 202, and temporary data for processing by the processor 201 is stored in the RAM 202. The ROM 203 includes, for example, a hard disk drive (HDD) and flash memory, which are electrically rewritable non-volatile memories. The ROM 203 stores programs and various types of data. The portable storage medium drive 204 is a device that reads data and programs stored in the portable storage medium 301. The portable storage medium 301 includes, for example, magnetic disks, optical discs, magneto-optical discs, and flash memory. The processor 201 executes programs stored in the ROM 203 or the portable storage medium 301 while cooperating with the RAM 202 or the ROM 203. Programs executed by the processor 201 and data to be accessed can be stored in another device capable of communicating with the computer.

[0119] The input / output device 205 is, for example, a keyboard, a touch screen, a display, etc. The input / output device 205 receives operation instructions such as from user operations and outputs the processing results of the computer.

[0120] In addition to a LAN card, etc., the communication interface 206 may also include, for example, a radio frequency receiver and a radio frequency transmitter, as well as an optical receiver and an optical transmitter.

[0121] These components of the computer are connected by a bus 207.

[0122] The ECU of the autonomous driving unit 11 installed in the truck 10 also includes at least components similar to the above-mentioned processor 201, RAM 202, and ROM 203. In the truck 10, communication with other ECUs and in-vehicle devices is performed through, for example, Controller Area Network (CAN), Local Interconnect Network (LIN), Ethernet (registered trademark), and FlexRay. Therefore, the communication interface provided in the ECU includes a CAN transceiver, etc., connected to the in-vehicle network.

[0123] In this embodiment, the server 20 includes an environment information acquisition unit 21, a data analysis unit 22, a classification unit 23, a map creation unit 24, an environment information storage unit 26, an analysis criterion storage unit 27, a static map storage unit 28, and a dynamic map storage unit 29, and the server 20 executes a map drawing process. However, the device for executing the map drawing process is not limited to the server 20. For example, these components can be provided in the truck 10, and the truck 10 can execute the map drawing process.

[0124] Those skilled in the art will readily understand that new embodiments can be achieved by omitting part of the technical concepts of various embodiments, freely combining part of the technical concepts of various embodiments, and substituting part of the technical concepts of various embodiments with known technologies.

[0125] List of Reference Numerals

[0126] 10 Truck

[0127] 20 Server

[0128] 30 Drone

[0129] 12 Driving Planning Unit

[0130] 13 Route Generation Unit

[0131] 14 Autopilot Control Unit

[0132] 16 Dynamic Map Storage Unit

[0133] 21 Environmental Information Acquisition Unit

[0134] 22 Data Analysis Unit

[0135] 23 Classification Unit

[0136] 24 Map Creation Unit

[0137] 26 Environmental Information Storage Unit

[0138] 27 Analysis Criterion Storage Unit

[0139] 28 Static Map Storage Unit

[0140] 29 Dynamic Map Storage Unit

[0141] 31 Camera

[0142] 32 Radar

[0143] 33 Lidar

Claims

1. A method for mapping a restricted area in which an autonomous vehicle drives, the method comprising the steps of: When the autonomous vehicle plans a new driving task, obtaining a first set of environmental information by a device flying above the restricted area including the driving starting point and the driving destination point of the autonomous vehicle, and obtaining a second set of environmental information after a time interval, each set of environmental information including the position and geometry of objects in the restricted area; Classifying objects present at the same position in the first set of environmental information and the second set of environmental information as semi-static objects, and classifying objects present at different positions in the first set of environmental information and the second set of environmental information as dynamic objects; And Creating a map including the position and geometry of each object classified as a semi-static object, wherein the first set of environmental information and the second set of environmental information are obtained in the step of obtaining the first set of environmental information and the second set of environmental information only when no most recent map has been created within a reference period determined based on the movement frequency of the semi-static objects.

2. The method for drawing a map according to claim 1, wherein, Determining the reference period based on the highest movement frequency obtained for each type of semi-static object in the restricted area.

3. The method for drawing a map according to claim 1 or 2, wherein, The map further includes the position and geometry of each object classified as a dynamic object.

4. The method for drawing a map according to claim 3, wherein, The map further includes the position and geometry of only those dynamic objects that move a distance less than or equal to a predetermined value in the first set of environmental information and the second set of environmental information.

5. The method for mapping according to claim 1 or 2, wherein When an object present at the same position in the first set of environmental information and the second set of environmental information is classified as a semi-static object, and if it is recognized based on the characteristics of the object that there is a possibility of the object moving, a map with a notification indicating the possibility of the object moving added thereto is created in the creating step.

6. The method for map drawing according to claim 1 or 2, wherein, In the classifying step, referring to a static map that pre-records the position and geometry of static objects that do not move, and classifying objects among the objects in the environmental information that are recorded in the static map as static objects.

7. The method for map drawing according to claim 1 or 2, wherein, In the step of obtaining the first set of environmental information and the second set of environmental information, obtaining the first set of environmental information and the second set of environmental information by at least one of a camera, a radar, and a lidar.

8. A mapping device configured to perform the steps according to any one of claims 1 to 7.

9. A vehicle equipped with the mapping device according to claim 8.

10. A computer program product including program code that, when run on a computer, is used to perform the steps according to any one of claims 1 to 7.

11. A computer-readable medium storing a computer program, the computer program including program code that, when run on a computer, is used to perform the steps according to any one of claims 1 to 7.

Images