Self-position estimation device
The self-position estimating device enhances position estimation accuracy by using laser radar to identify and exclude movable parts of dynamic objects from the comparison with map information, addressing the challenge of dynamic objects in autonomous driving.
Patent Information
- Application Number
- JP2021028396
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-02-25
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2041-02-25
AI Technical Summary
Existing self-position estimation technologies in autonomous driving face challenges in accurately estimating the position of a moving object when dynamic objects with movable parts are present, as these objects can distort the surrounding object information, leading to reduced robustness.
A self-position estimating device that utilizes a laser radar to acquire surrounding object information, stores fixed and movable part information of dynamic objects, and determines their presence, excluding the movable parts from the comparison with map information to enhance accuracy.
Improves position estimation accuracy by distinguishing between fixed and movable parts of dynamic objects, enabling precise self-position estimation even when dynamic objects are present.
Smart Images

Figure 0007581080000001 
Figure 0007581080000002 
Figure 0007581080000003
Abstract
Description
[Technical field]
[0001] The present invention relates to a self-location estimation device that estimates the position of a moving object. [Background technology]
[0002] Patent Document 1 describes a self-location determination device that measures its own location using a CCD camera and a laser pointer. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] Japanese Patent Application Publication No. 07-071978 Summary of the Invention [Problem to be solved by the invention]
[0004] In fields such as autonomous driving, self-location estimation is performed using technologies such as SLAM (Simultaneous Localization and Mapping). In SLAM, for example, objects present around a moving object are acquired as peripheral object information (laser point cloud) using LiDAR (Light Detection and Ranging), a type of laser radar, and the acquired peripheral object information is collated with map information (point cloud) to estimate the self-location of the moving object. However, if dynamic objects not included in the map information are present around the moving object, there is a possibility that the self-location of the moving object cannot be estimated by collation of the map information with the peripheral object information. In other words, there is a problem in that robustness against dynamic objects is low.
[0005] Here, it is possible to exclude the portion corresponding to the dynamic object from the peripheral object information and compare the map information with the peripheral object information. However, since a dynamic object such as a dump truck has a movable part whose shape changes, when the movable part moves, it becomes difficult to estimate the portion corresponding to the dynamic object in the peripheral object information.
[0006] Therefore, an object of the present invention is to provide a self-location estimation device that can improve the accuracy of location estimation when a dynamic object is present around a moving body. [Means for solving the problem]
[0007] The self-location estimation device of the present invention comprises a peripheral object information acquisition unit that acquires peripheral object information, which is information about objects present in the vicinity of a moving body; a memory unit in which map information and dynamic object information, which is information about dynamic objects, are stored; a dynamic object determination unit that determines whether or not a dynamic object is present in the vicinity of the moving body; and a self-location estimation unit that estimates the position of the moving body in the map information by matching the map information with the peripheral object information, wherein the dynamic object information has fixed portion information, which is information about fixed portions of the dynamic object whose form does not change, and movable portion information, which is information about movable portions of the dynamic object whose form changes, and the dynamic object determination unit determines whether or not a dynamic object is present in the vicinity of the moving body by matching the peripheral object information with the fixed portion information, and when the dynamic object determination unit determines that a dynamic object is present in the vicinity of the moving body, the self-location estimation unit excludes the portion corresponding to the dynamic object from the peripheral object information, and estimates the position of the moving body in the map information by matching the map information with the peripheral object information.
[0008] In this self-location estimation device, whether or not a dynamic object exists around the moving body is determined by comparing the peripheral object information with the fixed part information, so that even if the movable part of the dynamic object changes shape, whether or not a dynamic object exists around the moving body can be determined. Then, when it is determined that a dynamic object exists around the moving body, the part corresponding to the dynamic object is excluded from the peripheral object information, and the position of the moving body in the map information is estimated by comparing the map information with the peripheral object information, so that the accuracy of estimating the position of the moving body when a dynamic object exists around the moving body can be improved.
[0009] The storage unit may store a plurality of pieces of dynamic object information having the same fixed portion information and different movable portion information. In this self-location estimation device, the storage unit stores a plurality of pieces of dynamic object information having the same fixed portion information and different movable portion information, so that when the map information and the peripheral object information are referenced, a portion corresponding to the dynamic object can be excluded from the peripheral object information according to the form of the dynamic object. This can further improve the accuracy of estimating the position of the moving body when a dynamic object is present around the moving body.
[0010] The peripheral object information acquisition unit may be a laser radar. In this self-position estimation device, the peripheral object information acquisition unit is a laser radar, so that the peripheral object information can be acquired with high accuracy. Effect of the Invention
[0011] According to the present invention, it is possible to improve the accuracy of position estimation when a dynamic object is present around a moving body. [Brief description of the drawings]
[0012] [Figure 1] 1 is a block diagram showing a self-location estimation device according to an embodiment; [Diagram 2] 1A and 1B are schematic diagrams for explaining changes in the shape of a movable object. [Diagram 3] 3(a) and 3(b) are schematic diagrams showing examples of dynamic objects with different shapes of movable parts. [Figure 4] FIG. 1 is a schematic diagram showing an example in which a dynamic object exists around a moving object; [Diagram 5] FIG. 13 is a schematic diagram showing an example in which dynamic objects existing around a moving object are excluded. [Figure 6] 4 is a flowchart showing a processing operation of the self-location estimation device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0013] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In the following description, the same or corresponding elements are designated by the same reference numerals, and duplicated description will be omitted.
[0014] Fig. 1 is a block diagram showing a self-location estimation device according to an embodiment. As shown in Fig. 1, the self-location estimation device 1 according to this embodiment is a device mounted on a moving body 2 and for estimating the position of the moving body 2. The moving body 2 on which the self-location estimation device is mounted is not particularly limited, but examples thereof include automobiles and motorcycles. The self-location estimation device 1 includes a surrounding object information acquisition unit 3, a storage unit 4, and a control unit 5.
[0015] The peripheral object information acquisition unit 3 acquires peripheral object information, which is information on objects present around the moving body 2. The peripheral object information acquisition unit 3 may be, for example, a device such as a LiDAR. Laser A radar, a millimeter wave sensor, a camera, or the like can be used. In this embodiment, the surrounding object information acquisition unit 3 will be described as a LiDAR. The surrounding object information acquisition unit 3, which is a LiDAR, detects objects around the moving body 2 by using a laser beam. The surrounding object information acquisition unit 3 transmits a laser beam to the surroundings of the moving body 2 and receives the laser beam reflected by the objects, thereby detecting objects present around the moving body 2 as a laser point cloud (lidar point cloud). This laser point cloud becomes surrounding object information.
[0016] The memory unit 4 stores map information A and dynamic object information B. The map information A is information such as the positions and shapes of buildings and the positions and shapes of roads that are necessary for estimating the position of the moving body 2. The map information A is information that can be matched with surrounding object information, for example, point cloud information that can be matched with a laser point cloud acquired by the surrounding object information acquisition unit 3 as surrounding object information. The dynamic object information B is information on dynamic objects that may exist around the moving body 2. The dynamic object information B is information that can be matched with surrounding object information, for example, point cloud information that can be matched with a laser point cloud acquired by the surrounding object information acquisition unit 3 as surrounding object information.
[0017] As shown in Fig. 2, the dynamic object 10 has a fixed part 11 whose shape does not change, and a movable part 12 whose shape does change. For example, in the case of a dump truck, the vehicle body is the fixed part 11, and the loading platform that tilts relative to the vehicle body is the movable part 12. In the case of a crane truck, the vehicle body is the fixed part 11, and the crane part that tilts and expands and contracts relative to the vehicle body is the movable part 12. In the case of a backhoe, the vehicle body is the fixed part 11, and the backhoe part that tilts and bends relative to the vehicle body is the movable part 12.
[0018] Therefore, the dynamic object information B has fixed portion information b1, which is information on the fixed portion 11 of the dynamic object 10 whose form does not change, and movable portion information b2, which is information on the movable portion 12 of the dynamic object 10 whose form changes. The fixed portion information b1 and the movable portion information b2 are information that can be matched with peripheral object information, for example, point cloud information that can be matched with the laser point cloud acquired by the peripheral object information acquisition unit 3 as peripheral object information.
[0019] Further, the storage unit 4 stores a plurality of pieces of dynamic object information B having the same fixed part information b1 and different movable part information b2. For example, the storage unit 4 stores dynamic object information B, which is information on the dynamic object 10 shown in FIG. 3(a), and dynamic object information B, which is information on the dynamic object 10 shown in FIG. 3(b). In FIG. 3(a) and FIG. 3(b), the fixed part 11 of the dynamic object 10 is shown by a solid line, and the movable part 12 of the dynamic object 10 is shown by a dashed line. In FIG. 3(a) and FIG. 3(b), a dump truck is shown as the dynamic object 10, and the body of the dump truck is the fixed part 11, and the loading platform of the dump truck is the movable part 12. The dynamic object information B in FIG. 3(a) and the dynamic object information B in FIG. 3(b) have the same fixed part information b1, which is information on the body, which is the fixed part 11, but different movable part information b2, which is information on the loading platform, which is the movable part 12. Specifically, the movable part information b2 of the dynamic object information B shown in Figure 3(a) is information on the loading platform in a form that is not tilted relative to the vehicle body, and the movable part information b2 of the dynamic object information B shown in Figure 3(b) is information on the loading platform in a form that is tilted relative to the vehicle body.
[0020] The control unit 5 is an electronic control unit (ECU) having a CPU, a ROM, a RAM, etc. The control unit 5 executes various controls by loading a program stored in the ROM into the RAM and executing it with the CPU. The control unit 5 may be configured with a single electronic control unit, or may be configured with multiple electronic control units. The control unit 5 has a dynamic object determination unit 6 and a self-position estimation unit 7. The dynamic object determination unit 6 and the self-position estimation unit 7 may be configured with the same electronic control unit, or may be configured with different electronic control units.
[0021] The dynamic object determination unit 6 determines whether or not a dynamic object 10 exists around the moving object 2. That is, the dynamic object determination unit 6 determines whether or not a dynamic object 10 exists around the moving object 2 by comparing the peripheral object information acquired by the peripheral object information acquisition unit 3 with the fixed portion information b1 of the dynamic object information B stored in the storage unit 4. For example, the fixed portion information b1 of all the dynamic object information B stored in the storage unit 4 is compared with the peripheral object information acquired by the peripheral object information acquisition unit 3 by a known pattern matching technique or the like. Then, when a portion corresponding to at least one of the fixed portion information b1 is found in the peripheral object information, it is determined that the dynamic object 10 exists around the moving object 2. On the other hand, when a portion corresponding to any of the fixed portion information b1 is not found in the peripheral object information, it is determined that the dynamic object 10 does not exist around the moving object 2. Then, the dynamic object determination unit 6 transmits the determination result to the self-location estimation unit 7.
[0022] The self-position estimation unit 7 compares the map information A with the surrounding object information to estimate the position of the moving body 2 in the map information A. For example, the self-position estimation unit 7 compares the surrounding object information acquired by the surrounding object information acquisition unit 3 with the map information A stored in the storage unit 4 using a known pattern matching technique or the like to estimate the position of the moving body 2 in the map information A.
[0023] At this time, as shown in FIG. 4, if a dynamic object 10 exists around the moving object 2, it becomes difficult to estimate the position of the moving object 2 in the map information A by matching the map information A with the peripheral object information. Therefore, when the dynamic object determination unit 6 determines that the dynamic object 10 exists around the moving object 2, the self-location estimation unit 7, as shown in FIG. 5, excludes a portion corresponding to the dynamic object 10 from the peripheral object information and estimates the position of the moving object 2 in the map information A by matching the map information A with the peripheral object information. In other words, when the dynamic object determination unit 6 determines that the dynamic object 10 exists around the moving object 2, it excludes a portion corresponding to both the fixed part 11 and the movable part 12 of the dynamic object 10 from the peripheral object information and estimates the position of the moving object 2 in the map information A by matching the map information A with the peripheral object information. Such estimation can be performed, for example, by generating matching peripheral object information in which the portion corresponding to the dynamic object 10 is removed from the peripheral object information and matching the generated matching peripheral object information with the map information A.
[0024] Next, an example of the processing operation of the self-position estimation device 1 will be described with reference to FIG.
[0025] First, the surrounding object information acquisition unit 3 acquires surrounding object information, which is information about objects existing around the moving object 2 (step S1).
[0026] Next, the dynamic object determination unit 6 compares the surrounding object information acquired by the surrounding object information acquisition unit 3 with the fixed portion information b1 of the dynamic object information B stored in the memory unit 4 (step S2), and determines whether or not a dynamic object 10 exists in the vicinity of the moving body 2 (step S3).
[0027] If it is determined that no dynamic object 10 exists around the moving object 2 (step S3: NO), the self-position estimation unit 7 estimates the position of the moving object 2 in the map information A by matching the map information A with the surrounding object information (step S5).
[0028] On the other hand, when it is determined that a dynamic object 10 is present in the vicinity of the moving body 2 (step S3: YES), the self-position estimation unit 7 excludes the portion corresponding to the dynamic object 10 from the surrounding object information (step S4), and estimates the position of the moving body 2 in the map information A by comparing the map information A with the surrounding object information (step S5).
[0029] As described above, in the self-location estimation device 1 according to the present embodiment, whether or not the dynamic object 10 exists in the vicinity of the moving body 2 is determined by comparing the peripheral object information with the fixed portion information b1, so that even if the movable portion 12 of the dynamic object 10 changes shape, whether or not the dynamic object 10 exists in the vicinity of the moving body 2 can be determined. Then, when it is determined that the dynamic object 10 exists in the vicinity of the moving body 2, the portion corresponding to the dynamic object 10 is excluded from the peripheral object information, and the position of the moving body 2 in the map information A is estimated by comparing the map information A with the peripheral object information, so that the accuracy of estimating the position of the moving body 2 when the dynamic object 10 exists in the vicinity of the moving body 2 can be improved.
[0030] Furthermore, since the storage unit 4 stores multiple pieces of dynamic object information B having the same fixed part information b1 and different movable part information b2, when comparing the map information A with the peripheral object information, it is possible to exclude a portion corresponding to the dynamic object 10 from the peripheral object information in accordance with the form of the dynamic object 10. This makes it possible to further improve the accuracy of estimating the position of the moving object 2 when the dynamic object 10 is present in the vicinity of the moving object 2.
[0031] Furthermore, since the peripheral object information acquisition unit 3 is a laser radar, peripheral object information can be acquired with high accuracy.
[0032] Although the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and may be modified or applied to other things without changing the gist of the claims. For example, in the above embodiment, it has been described that two pieces of dynamic object information having the same fixed part information and different movable part information are stored in the storage unit, but in cases where the shape change of the movable part is complex, three or more pieces of dynamic object information may be stored in the storage unit. [Explanation of symbols]
[0033] 1...self-location estimation device, 2...moving body, 3...surrounding object information acquisition unit, 4...memory unit, 5...control unit, 6...dynamic object determination unit, 7...self-location estimation unit, 10...dynamic object, 11...fixed part, 12...movable part, A...map information, B...dynamic object information, b1...fixed part information, b2...movable part information.
Claims
1. a surrounding object information acquisition unit that acquires surrounding object information, which is information about objects present around the moving object; A storage unit in which map information and dynamic object information that is information on dynamic objects are stored; a dynamic object determination unit that determines whether or not the dynamic object is present in the vicinity of the moving object; a self-location estimation unit that estimates a location of the moving object in the map information by comparing the map information with the peripheral object information, the dynamic object information includes fixed portion information that is information on a fixed portion of the dynamic object whose shape does not change, and movable portion information that is information on a movable portion of the dynamic object whose shape changes, the dynamic object determination unit determines whether or not the dynamic object is present in the vicinity of the moving object by comparing the peripheral object information with the fixed portion information; When the dynamic object determination unit determines that the dynamic object is present around the moving object, the self-location estimation unit estimates a position of the moving object in the map information by comparing the map information with peripheral object information obtained by excluding a portion corresponding to the dynamic object from the peripheral object information. Self-location estimation device.
2. the storage unit stores a plurality of pieces of dynamic object information, the fixed part information being the same and the movable part information being different; The self-location estimation device according to claim 1 .
3. The peripheral object information acquisition unit is a laser radar. The self-location estimation device according to claim 1 or 2.
Citation Information
Patent Citations
One's own position recognition system
JP1995071978A
Position estimating device, control method, program, and storage medium
JP2020034452A
Map generator, control method, program, and storage medium
JP2020034771A
Stationary object data generator, method for control, program, and storage medium
JP2020034772A
Moving object position estimating device and moving object position estimating program
JP2020125960A
