Vehicle parking method and system combining obstacle points and obstacle fusion boxes

Abstract

This invention discloses a vehicle parking method and system that combines obstacle points and obstacle fusion boxes. Based on the DBSCAN algorithm, obstacle points detected by the vehicle's radar are clustered into clusters, and an obstacle fusion box is fitted to these clusters using a convex hull algorithm. The nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body are calculated respectively. Obstacle points and obstacle fusion boxes whose nearest distance to the vehicle body is greater than a threshold obstacle distance are deleted, resulting in filtered real obstacle points and real obstacle fusion boxes. The vehicle is then parked based on these real obstacle points and real obstacle fusion boxes. This not only filters out noise points but also identifies the accurate location and range of large obstacles by fitting the obstacle fusion boxes, providing a realistic and accurate obstacle location reference for automatic parking and improving the accuracy of automatic parking.

Description

Technical Field

[0001] This invention relates to automatic parking technology for vehicles, and more particularly to a vehicle parking method and system that combines obstacle points and obstacle fusion frames. Background Technology

[0002] Automatic parking technology often uses ultrasonic radar to detect the distance and coordinates of obstacles, enabling the vehicle to avoid them during parking and achieve a safer parking experience. However, in each detection cycle, an ultrasonic radar sensor can only detect one obstacle point. Even if the obstacle is a wall or pillar, the radar sensor can only detect a series of scattered obstacle points during parking. These scattered obstacle points are not enough to accurately identify the specific location and size range of the obstacle. This does not significantly improve the effectiveness of obstacle avoidance and the selection of the best parking space, resulting in low parking accuracy. Summary of the Invention

[0003] The purpose of this invention is to overcome the above-mentioned technical deficiencies and propose a vehicle parking method and system that combines obstacle points and obstacle fusion frames, thereby solving the problem of low obstacle recognition accuracy caused by the scattered obstacle points of existing ultrasonic radar.

[0004] To achieve the above-mentioned technical objectives, the first aspect of the present invention provides a vehicle parking method combining obstacle points and obstacle fusion frames, which includes the following steps:

[0005] Obstacle points around the vehicle are obtained by cyclically detecting the vehicle's body radar.

[0006] Based on the DBSCAN algorithm, obstacle points are clustered into clusters using Euclidean distance relationships;

[0007] The obstacle fusion box is fitted by clustering obstacle points using the convex hull algorithm.

[0008] Calculate the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body, respectively. Delete obstacle points and obstacle fusion boxes whose nearest distance to the vehicle body is greater than the threshold obstacle distance, and obtain the filtered real obstacle points and real obstacle fusion boxes.

[0009] Park the vehicle based on the real obstacle points and the real obstacle fusion box.

[0010] A second aspect of the present invention provides a vehicle parking system that combines obstacle points and obstacle fusion frames, comprising the following functional modules:

[0011] The radar detection module is used to cyclically detect obstacle points around the vehicle using the vehicle's radar.

[0012] The obstacle point clustering module is used to cluster obstacle points into clusters based on the DBSCAN algorithm using Euclidean distance.

[0013] The obstacle point fusion module is used to fit the obstacle point clusters to generate the obstacle fusion box using the convex hull algorithm;

[0014] The obstacle filtering module is used to calculate the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body, respectively. It deletes obstacle points and obstacle fusion boxes whose nearest distance from the vehicle body is greater than the threshold obstacle distance, and obtains the filtered real obstacle points and real obstacle fusion boxes.

[0015] The parking module is used to park the vehicle based on the real obstacle points and the real obstacle fusion box.

[0016] A third aspect of the present invention provides a server including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the above-described vehicle parking method combining obstacle points and obstacle fusion boxes.

[0017] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described vehicle parking method combining obstacle points and obstacle fusion frames.

[0018] Compared with existing technologies, the vehicle parking method and system combining obstacle points and obstacle fusion boxes described in this invention uses the DBSCAN algorithm to cluster obstacle points obtained by the vehicle's radar through cyclic detection into clusters, and then uses the convex hull algorithm to fit obstacle point clusters into obstacle fusion boxes. It calculates the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body, respectively, and deletes obstacle points and obstacle fusion boxes whose nearest distance to the vehicle body is greater than a threshold obstacle distance, obtaining filtered real obstacle points and real obstacle fusion boxes. The vehicle is then parked based on these real obstacle points and real obstacle fusion boxes. This not only filters out noise points but also identifies the accurate location and range of large obstacles by fitting obstacle fusion boxes, providing accurate obstacle position references for automatic vehicle parking and improving the accuracy of automatic parking. Furthermore, it is not limited by lighting conditions and can be implemented under any lighting conditions. Attached Figure Description

[0019] Figure 1 This is a flowchart of the vehicle parking method combining obstacle points and obstacle fusion boxes according to an embodiment of the present invention;

[0020] Figure 2 This is a schematic diagram of the ultrasonic radar detection area described in an embodiment of the present invention;

[0021] Figure 3 This is a block diagram of a vehicle parking system combining obstacle points and obstacle fusion frames as described in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0023] like Figure 1 As shown, an embodiment of the present invention provides a vehicle parking method that combines obstacle points and obstacle fusion boxes, which includes the following steps:

[0024] S1. Obstacle points around the vehicle are obtained by using vehicle radar to cyclically detect them.

[0025] Multiple ultrasonic radars are installed on the vehicle body. The detection area of ​​the vehicle body radar needs to be at least around the vehicle body. Each probe of the ultrasonic radar can only detect the distance and coordinate information of the nearest obstacle in each detection cycle. For example, according to the needs of the vehicle body length and width, 12 ultrasonic radars are installed around the vehicle body. Then, in one detection cycle, the 12 ultrasonic radar probes can provide the distance and coordinate information of 12 obstacles around the vehicle.

[0026] Because ultrasonic radar can only detect a fan-shaped area, the detection error is significant in areas that are too close or too far from the radar. Therefore, during the process of using vehicle-mounted radar to cyclically detect obstacles around the vehicle, obstacles outside the radar's effective detection area are deleted. Specifically, as follows... Figure 2 As shown, the dashed line represents the edge of the area detectable by the ultrasonic radar, and the area within the two solid arcs is the effective detection area of ​​the ultrasonic radar. If the obstacle coordinates provided by the ultrasonic radar are outside the range of the dashed line or outside the effective range, then this obstacle point is invalid and should be deleted.

[0027] In addition, after using vehicle radar to cyclically detect obstacles around the vehicle, it is also necessary to filter these obstacles.

[0028] S2. Based on the DBSCAN algorithm, obstacle points are clustered into clusters using Euclidean distance.

[0029] S3. Use the convex hull algorithm to cluster the obstacle points and fit the obstacle fusion box.

[0030] To update and maintain obstacle information in real time, this method maintains an obstacle point list dataset and an obstacle fusion box list dataset.

[0031] When an obstacle point is detected in the first loop, the obstacle point needs to be clustered into clusters based on the Euclidean distance relationship using the DBSCAN algorithm, and then the obstacle fusion box is fitted to the obstacle point clusters using the convex hull algorithm.

[0032] The obstacle points detected by the radar in the first loop are directly added to the obstacle point list dataset. The DBSCAN algorithm is then used to cluster all the initially detected obstacle points into clusters based on Euclidean distance. If obstacle point clustering fails, the point remains in the obstacle point list. If obstacle point clustering is successful, the convex hull algorithm is used to calculate the bounding rectangle of the cluster, i.e., the obstacle fusion box. If an obstacle fusion box can be obtained, the information of the obstacle points in this cluster and the obtained obstacle fusion box information are added to the obstacle fusion box list, and the information of the obstacle points in this cluster is removed from the obstacle point list.

[0033] When a new obstacle point is detected in the second or subsequent iterations, it is only necessary to merge the new obstacle point with the existing obstacle fusion box, and then use the convex hull algorithm to fit the obstacle point clusters to obtain the obstacle fusion box.

[0034] When a new obstacle point is detected by the radar in the second or subsequent cycles, the filtered new obstacle point is fused with the existing obstacle fusion box. That is, based on the coordinates of the new obstacle point and the coordinate range of the existing obstacle fusion box, it is determined whether the new obstacle point falls within the range of any existing obstacle fusion box.

[0035] If a new obstacle point falls within the range of any existing obstacle merging box, it is added to the point set of that obstacle merging box, and the merging is successful.

[0036] If the new obstacle point does not fall within the merge box of any existing obstacle, the merging of the obstacle point is considered to have failed.

[0037] If a new obstacle point cannot be integrated into any obstacle fusion box, it is compared with the existing obstacle points one by one to determine if they are the same point. If the distance between the new obstacle point and one of the obstacle points in the obstacle point list is less than 5cm, then the two obstacle points are considered to be the same point. The coordinates of the obstacle point in the obstacle point list are then updated to the coordinates of the new obstacle point, and the successful tracking count of the obstacle point is incremented by 1.

[0038] If a new obstacle point cannot be merged into the obstacle merge box and does not belong to the original obstacle points, then this new obstacle point is considered a newly added obstacle point and is added to the obstacle point list.

[0039] S4. Calculate the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body respectively. Delete the obstacle points and obstacle fusion boxes whose nearest distance from the vehicle body is greater than the threshold obstacle distance, and obtain the filtered real obstacle points and real obstacle fusion boxes.

[0040] Since obstacle detection occurs during automatic parking, the vehicle may approach or move away from previously detected obstacles while parking. The system calculates the distance between points in the obstacle point list and the vehicle body. If the distance is greater than the safe parking distance (e.g., 6 meters), the obstacle point is considered to pose no safety threat to parking and is removed from the obstacle point list. Similarly, the system calculates the closest distance between the coordinates of the bounding boxes in the obstacle fusion box list and the vehicle body. If the closest distance between the bounding box and the vehicle body is greater than 6 meters, all obstacles within the bounding box are considered to pose no safety threat to parking, and the bounding box and all obstacles within it are removed from the obstacle fusion box list.

[0041] Because ultrasonic radar can detect obstacles with false detections, if an obstacle can be tracked multiple times, it indicates that the probability of a false detection is extremely low. Therefore, before calculating the closest distance, the number of times an obstacle is successfully tracked is first counted. That is, if an obstacle is successfully tracked less than once within a preset detection period, it is considered a noise point and removed from the obstacle list.

[0042] Finally, the remaining obstacle point information in the obstacle point list and the remaining obstacle fusion box information in the obstacle fusion box list represent the real obstacle points and real obstacle fusion boxes around the vehicle at the current moment.

[0043] S5. Park the vehicle based on the actual obstacle points and the actual obstacle fusion box.

[0044] This invention uses the DBSCAN algorithm to cluster obstacle points detected by the vehicle's radar into clusters, and then uses the convex hull algorithm to fit obstacle point fusion boxes to these clusters. The invention calculates the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body, respectively. Obstacle points and fusion boxes whose nearest distance to the vehicle body is greater than a threshold obstacle distance are deleted, resulting in filtered real obstacle points and real obstacle fusion boxes. The vehicle is then parked based on these real obstacle points and fusion boxes. This vehicle parking method combining obstacle points and fusion boxes can filter out noise points and identify the accurate location and range of large obstacles by fitting fusion boxes, thus providing a realistic and accurate obstacle position reference for automatic parking and improving the accuracy of automatic parking. Furthermore, it is not limited by lighting conditions and can be implemented under any lighting conditions.

[0045] like Figure 3 As shown, this embodiment of the invention also discloses a vehicle parking system that combines obstacle points and obstacle fusion boxes, which includes the following functional modules:

[0046] Radar detection module 10 is used to obtain obstacle points around the vehicle by using vehicle body radar to cyclically detect them;

[0047] The obstacle point clustering module 20 is used to cluster obstacle points into clusters based on the DBSCAN algorithm using Euclidean distance.

[0048] The obstacle point fusion module 30 is used to fit the obstacle point clusters to generate the obstacle fusion box using the convex hull algorithm;

[0049] The obstacle filtering module 40 is used to calculate the closest distance between the obstacle point and the vehicle body, and the closest distance between the obstacle fusion box and the vehicle body, respectively, and delete the obstacle points and obstacle fusion boxes that are closer to the vehicle body than the threshold obstacle distance, so as to obtain the filtered real obstacle points and real obstacle fusion boxes.

[0050] The parking module 50 is used to park the vehicle based on the real obstacle points and the real obstacle fusion box.

[0051] The execution method of the vehicle parking system combining obstacle points and obstacle fusion boxes in this embodiment is basically the same as the vehicle parking method combining obstacle points and obstacle fusion boxes described above, so it will not be described in detail.

[0052] In this embodiment, the server is a device that provides computing services, typically referring to a computer with high computing power that is provided to multiple consumers via a network. The server in this embodiment includes a memory, a processor, and a system bus. The memory includes executable programs stored thereon. Those skilled in the art will understand that the terminal device structure of this embodiment does not constitute a limitation on the terminal device and may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0053] The memory can be used to store software programs and modules. The processor executes various terminal functions and data processing by running the software programs and modules stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, application programs required for at least one function (such as sound playback, image playback, etc.), etc.; the data storage area can store data created according to the use of the terminal (such as audio data, phonebook, etc.). In addition, the memory can include high-speed random access memory, and can also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0054] An executable program containing a vehicle parking method combining obstacle points and obstacle fusion frames is stored in memory. This executable program can be divided into one or more modules / units, which are stored in the memory and executed by a processor to complete the information acquisition and implementation process. Each module / unit can be a series of computer program instruction segments capable of performing a specific function, describing the execution process of the computer program on the server. For example, the computer program can be divided into a radar detection module 10, an obstacle point clustering module 20, an obstacle point fusion module 30, an obstacle filtering module 40, and a parking module 50.

[0055] The processor is the control center of the server, connecting various parts of the terminal device through various interfaces and lines. It performs various terminal functions and processes data by running or executing software programs and / or modules stored in memory, and by calling data stored in memory, thereby providing overall monitoring of the terminal. Optionally, the processor may include one or more processing units; preferably, the processor may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, applications, etc., and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into the processor.

[0056] The system bus connects various functional components within a computer, transmitting data, address, and control information. Types of system buses include PCI, ISA, and VESA. Processor instructions are transmitted to memory via the bus, and memory sends data back to the processor. The system bus handles the data and instruction exchange between the processor and memory. Of course, the system bus can also connect to other devices, such as network interfaces and display devices.

[0057] The server should include at least a CPU, chipset, memory, and disk system; other components will not be described in detail here.

[0058] In this embodiment of the invention, the executable program executed by the processor included in the terminal is specifically: a vehicle parking method combining obstacle points and obstacle fusion boxes, which includes the following steps:

[0059] Obstacle points around the vehicle are obtained by cyclically detecting the vehicle's body radar.

[0060] Based on the DBSCAN algorithm, obstacle points are clustered into clusters using Euclidean distance relationships;

[0061] The obstacle fusion box is fitted by clustering obstacle points using the convex hull algorithm.

[0062] Calculate the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body, respectively. Delete obstacle points and obstacle fusion boxes whose nearest distance to the vehicle body is greater than the threshold obstacle distance, and obtain the filtered real obstacle points and real obstacle fusion boxes.

[0063] Park the vehicle based on the real obstacle points and the real obstacle fusion box.

[0064] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0065] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0066] Those skilled in the art will recognize that the modules, units, and / or method steps of the various embodiments described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of the invention.

[0067] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A vehicle parking method combining obstacle points and obstacle fusion frames, characterized in that, Includes the following steps: Obstacle points around the vehicle are obtained by cyclically detecting the vehicle's body radar. Based on the DBSCAN algorithm, obstacle points are clustered into clusters using Euclidean distance relationships; The obstacle fusion box is fitted by clustering obstacle points using the convex hull algorithm. Calculate the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body, respectively. Delete obstacle points and obstacle fusion boxes whose nearest distance to the vehicle body is greater than the threshold obstacle distance, and obtain the filtered real obstacle points and real obstacle fusion boxes. Park the vehicle based on the real obstacle points and the real obstacle fusion box; When an obstacle point is detected in the first loop, the obstacle point needs to be clustered into clusters based on the Euclidean distance relationship using the DBSCAN algorithm, and then the obstacle fusion box is fitted to the obstacle point clusters using the convex hull algorithm. When a new obstacle point is detected in the second or subsequent iterations, it is only necessary to merge the new obstacle point with the existing obstacle fusion box, and then use the convex hull algorithm to fit the obstacle point clusters to obtain the obstacle fusion box. The process of fusing the new obstacle point with the existing obstacle fusion box specifically includes: Based on the coordinates of the obstacle point and the coordinate range of the existing obstacle merge box, determine whether the new obstacle point falls within the range of any existing obstacle merge box; If a new obstacle point falls within the range of any existing obstacle merging box, it is added to the point set of that obstacle merging box, and the merging is successful. If the new obstacle point does not fall within the merge box of any existing obstacle, the merging of the obstacle point is considered to have failed.

2. The vehicle parking method combining obstacle points and obstacle fusion frames according to claim 1, characterized in that, During the process of using vehicle radar to cyclically detect obstacles around the vehicle, obstacles that are outside the effective detection area of ​​the radar are deleted; after using vehicle radar to cyclically detect obstacles around the vehicle, the obstacles need to be filtered.

3. The vehicle parking method combining obstacle points and obstacle fusion frames according to claim 1, characterized in that, If a new obstacle point cannot be merged into any obstacle merging box, it is compared with the existing obstacle points one by one to determine if they are the same point. If they are the same point, the obstacle point is updated and the successful tracking count of the obstacle point is incremented by 1.

4. The vehicle parking method combining obstacle points and obstacle fusion frames according to claim 1, characterized in that, If a new obstacle point cannot be merged into the obstacle merging box, nor does it belong to the original obstacle points, then this new obstacle point is considered a newly added obstacle point.

5. The vehicle parking method combining obstacle points and obstacle fusion frames according to claim 3, characterized in that, If an obstacle point is successfully tracked less than once within a preset detection period, it will be evaluated as a noise point and deleted.

6. A vehicle parking system combining obstacle points and obstacle fusion frames, characterized in that, To implement the method as described in any one of claims 1-5, the method comprises the following functional modules: The radar detection module is used to cyclically detect obstacle points around the vehicle using the vehicle's radar. The obstacle point clustering module is used to cluster obstacle points into clusters based on the DBSCAN algorithm using Euclidean distance. The obstacle point fusion module is used to fit the obstacle point clusters to generate the obstacle fusion box using the convex hull algorithm; The obstacle filtering module is used to calculate the nearest distance between the obstacle point and the vehicle body, and the nearest distance between the obstacle fusion box and the vehicle body, respectively. It deletes obstacle points and obstacle fusion boxes whose nearest distance from the vehicle body is greater than the threshold obstacle distance, and obtains the filtered real obstacle points and real obstacle fusion boxes. The parking module is used to park the vehicle based on the real obstacle points and the real obstacle fusion box.

7. A server comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the vehicle parking method combining obstacle points and obstacle fusion frames as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the vehicle parking method combining obstacle points and obstacle fusion frames as described in any one of claims 1 to 5.

Images