Method and system for optimizing association between target and road boundary
By calculating the number and location of the road boundary radar reflection points around the target object and determining and limiting their correlation distance, the error correlation problem of the AEB system when identifying stationary vehicles is solved, and the accuracy of obstacle detection is improved.
Patent Information
- Application Number
- CN202310582427.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-22
- Publication Date
- 2024-11-22
AI Technical Summary
Existing AEB systems tend to incorrectly associate stationary vehicles to road boundaries when identifying stationary vehicles, affecting the accuracy of automatic emergency braking, especially when the vehicle is parked near the road boundary, resulting in a decrease in the accuracy of obstacle detection.
By calculating the number and location of radar reflection points around the target object, determine whether the target is close to the road boundary, and limit its correlation distance according to the threshold to prevent incorrect correlation. The specific steps include the first determining step: judging the target to the road boundary, the second determining step: determining which side the road boundary is close to, and the association restriction step: limit the association based on distance.
Effectively prevent road boundaries from being erroneously associated with targets, improving the accuracy of obstacle detection, especially for the identification and classification of static obstacles.
Smart Images

Figure CN118991779A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent driving technology, and in particular to a method for optimizing the association relationship between a target and a road boundary and a system for optimizing the association relationship between a target and a road boundary. Background Art
[0002] As a major basic component of ADAS (Advanced Driver Assistance System), the main purpose of AEB (Automatic Emergency Braking) is to be able to select the vehicle in front of the autonomous driving lane as the primary target, and to control the longitudinal motion state and subsequent distance of the vehicle in real time and comfortably according to the motion state of the primary target.
[0003] For 1R systems (1R systems refer to assisted driving function systems that only use radar as a sensor to detect targets), it is difficult to identify stationary objects because radar is more sensitive to moving objects than stationary objects. Moreover, existing systems are designed to identify and select objects on the street (such as vehicles, bicycles and pedestrians), so they generally tend to optimize more on these objects. However, stationary vehicles are also objects on the street, and they should also be selected by AEB as obstacles. It can be seen that when the target vehicle is parked near the road boundary, it can easily be mistakenly associated with the road boundary, which will affect AEB's selection of it. On the other hand, the location of the road boundary will also affect the detection of the target. Summary of the invention
[0004] The present invention aims to provide a method for optimizing the association relationship between a target and a road boundary and a system for optimizing the association relationship between a target and a road boundary to improve the obstacle detection accuracy.
[0005] A method for optimizing the association relationship between a target and a road boundary according to one aspect of the present invention comprises:
[0006] A first determining step of determining that the target is close to a road boundary based on the number of road boundaries within a predetermined distance around the target;
[0007] A second determination step of determining whether the target is close to a left road boundary or a right road boundary; and
[0008] The association restriction step restricts whether the object is associated with the left road boundary or the right road boundary based on the distance between the object and the left road boundary or the right road boundary.
[0009] A system for optimizing the association relationship between a target and a road boundary according to one aspect of the present invention comprises:
[0010] A first determination module determines that the target is close to a road boundary based on the number of road boundaries within a predetermined distance around the target;
[0011] A second determination module determines whether the target is close to a left road boundary or a right road boundary; and
[0012] The association restriction module restricts whether the object is associated with the left or right road boundary based on the distance between the object and the left or right road boundary.
[0013] A computer-readable medium according to one aspect of the present invention stores a computer program thereon, and when the computer program is executed by a processor, the method for optimizing the association relationship between a target and a road boundary is implemented.
[0014] A computer device according to one aspect of the present invention comprises a storage module, a processor and a computer program stored in the storage module and executable on the processor, wherein the processor implements the method for optimizing the association relationship between a target and a road boundary when executing the computer program.
[0015] An automatic emergency braking system according to one aspect of the present invention is characterized in that it includes the system for optimizing the association relationship between the target and the road boundary. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] FIG1( a ) is a schematic diagram showing that the road boundary position is incorrectly associated with an object.
[0017] FIG1( b ) is a schematic diagram showing the correlation relationship between the target and the road boundary after the optimization according to the present invention.
[0018] Figure 2 Detailed description of the invention is provided for a flow chart of a method for optimizing the association between a target and a road boundary.
[0019] Figure 3 is a schematic diagram showing an example of the first determination step.
[0020] Figure 4 is a schematic diagram showing an example of the association restriction step.
[0021] Figure 5 It is a structural block diagram of a system for optimizing the association relationship between a target and a road boundary of the present invention. DETAILED DESCRIPTION
[0022] The following describes some of the various embodiments of the present invention, which are intended to provide a basic understanding of the present invention, but are not intended to identify the key or decisive elements of the present invention or to limit the scope of protection.
[0023] For the purpose of brevity and illustration, the principles of the present invention are described herein mainly with reference to its exemplary embodiments. However, those skilled in the art will readily recognize that the same principles are equally applicable to and can be implemented in all types of methods for optimizing the association relationship between a target and a road boundary and systems for optimizing the association relationship between a target and a road boundary, and any such changes do not depart from the true spirit and scope of the present patent application.
[0024] First, the technical concept of the present invention will be described.
[0025] The principle of target recognition and tracking by millimeter wave is as follows: after the receiving antenna receives the radar echo and demodulates it, the controller digitally samples the analog signal and performs corresponding filtering, then transforms the signal into the frequency domain, and then searches for the distance, speed and angle information in the signal. At this step, the required target cannot be found, and only the information of the reflection point of the radar wave is obtained. For many high-performance millimeter wave radars, the multiple reflection points obtained at this time may come from one object, for example, a truck may form 5 to 10 reflection points. Therefore, first of all, the reflection points belonging to the same object must be matched to a reflection point cluster, and then by tracking each reflection point cluster, a guess about the distribution of the object is formed. In the next measurement cycle, based on the object distribution of the previous measurement cycle, the possible object distribution in this measurement cycle is predicted, and then the current reflection point cluster is tried to be matched with the predicted result, for example, by comparing the parameters such as the position and speed of the object. When the reflection point cluster successfully matches the object information obtained in the previous test cycle, the trajectory of the object is obtained.
[0026] By studying the above working principle of millimeter wave, the inventor of the present invention found that: since the millimeter wave radar perception depends on the position point of the reflected wave of the target reflector, when a certain threshold is reached, the position points on different reflectors will be clustered into different targets. When it comes to updating the target object, the system will compare the radial distance and radial velocity between the object and the position. If the position can pass the above two thresholds, it will be associated with the target object. However, since the vehicle is parked on the side of the road, the position of the road boundary will be seriously misassociated with the object, as shown in Figure 1(a). Figure 1(a) is a schematic diagram showing that the road boundary position is incorrectly associated with the object. The x-axis in Figure 1(a) is the longitudinal axis, the y-axis is the transverse axis (radial), and the origin is the center of the rear axle of the vehicle. In Figure 1(a), the circled "No. 72" radar point should have been a radar point from the road boundary (RSB), but it was mistakenly associated with the real vehicle target A. The dotted line in Figure 1(a) indicates this wrong association.
[0027] Based on the above findings, the method for optimizing the association relationship between a target and a road boundary of the present invention is proposed. The main technical concept of the present invention is to first determine whether the target in front should be optimized. For example, the number of SEPs (stationary environment processing, indicating static targets detected by radar) around the target object can be calculated. If the count value exceeds the threshold, the target object is marked as needing to be optimized. In addition, a flag indicating whether the target object exists on the right or left side is further provided. This flag will be used for updating the target object. Furthermore, when updating the target object, if the target object exists on the right side, the association distance with the road boundary on the right side is limited. If the target object exists on the left side, the association distance with the road boundary on the left side is limited. This updating process can be implemented, for example, by PER (perception module).
[0028] Next, a method for optimizing the association relationship between the target and the road boundary according to the present invention will be described.
[0029] Figure 2 Detailed description of the invention is provided for a flow chart of a method for optimizing the association between a target and a road boundary.
[0030] like Figure 2 As shown, the method for optimizing the association relationship between the target and the road boundary of the present invention includes:
[0031] First determination step S100: determining that the target is close to a road boundary based on the number of road boundaries within a predetermined distance around the target;
[0032] Second determination step S200: determining whether the target is close to the left road boundary or the right road boundary; and
[0033] The association restriction step S300: restricting whether the target is associated with the left or right road boundary based on the distance between the target and the left or right road boundary.
[0034] The first determination step S100 includes:
[0035] Obtain radar reflection points of the target and the road boundary;
[0036] Determine the target as a static target;
[0037] Counting the number of radar reflection points of a road boundary within a predetermined distance around the radar reflection point of the target; and
[0038] When the number of radar reflection points of the road boundary within a predetermined distance around the radar reflection point of the target exceeds a preset first threshold, it is determined that the target is close to the road boundary.
[0039] Among them, if the number of radar reflection points at the left road boundary exceeds a preset first threshold, then in the second determination step S200, it is determined that the target is close to the left road boundary; if the number of radar reflection points at the right road boundary exceeds the preset first threshold, then in the second determination step S200, it is determined that the target is close to the right road boundary.
[0040] As an example, see Figure 3 The first determination step is described. Figure 3 is a schematic diagram showing an example of the first determination step.
[0041] like Figure 3 As shown, the first determination step includes the following sub-steps:
[0042] S10: Start;
[0043] S11: traverse each static target on the left and right sides;
[0044] S12: traverse each road boundary object;
[0045] S13: Determine whether the road boundary object is within the predetermined distance of the static target. If the determination result is yes (Y), proceed to S14; otherwise (N), proceed to S16;
[0046] S14: Determine whether the number of road boundaries around the target reaches a preset first threshold. If the determination result is yes (Y), proceed to S15; otherwise (N), proceed to S16;
[0047] S15: determining that the current target appears beside a road boundary, that is, determining that the current target is close to the road boundary;
[0048] S16: End.
[0049] Here, let's give another specific example. For an object that is within 1.5 meters from the y-axis and less than 50 meters from the x-axis, a window is used to calculate the number of road boundaries next to the object. Iterate each target, and then calculate the distance difference on the y-axis. If the distance difference is less than 5 meters and the distance difference on the x-axis is also less than 20 meters, it is identified as a valid object and 1 is added to the counter. In addition, two values (right side count value and right side count value) are calculated from the right and left sides respectively. If one of them is greater than the first threshold (for example, 5 times), the target is marked as close to the road boundary. The specific value used as the first threshold here can be adjusted and set according to reference factors such as the road, road boundary, the vehicle itself, and practical experience. The specific value is not limited here.
[0050] That is to say, when the left side count value exceeds the first threshold value in the first determination step S100, the target is determined to be the road boundary close to the left side in the second determination step S200; when the right side count value exceeds the first threshold value in the first determination step S100, the target is determined to be the road boundary close to the right side in the second determination step S200.
[0051] Next, the association restriction step S300 will be described.
[0052] When it is determined in the second determination step S200 that the target is close to the left road boundary, then in the association restriction step S300 when the radial distance between the target and the left road boundary is greater than a preset second threshold value, the target is restricted to be associated with the left road boundary; when it is determined in the second determination step S200 that the target is close to the right road boundary, then in the association restriction step S300 when the radial distance between the target and the right road boundary is greater than a preset second threshold value, the target is restricted to be associated with the right road boundary.
[0053] Figure 4 is a schematic diagram showing an example of an association restriction step. Specifically, Figure 4 It means limiting the associated distance of the target object when updating the target. Here, the target is the detected radar point, and the associated distance refers to the associated distance between the target and the radar point, that is, judging whether the radar point originates from the target.
[0054] like Figure 4 As shown, when updating the target, limiting the association distance of the target object in the association limitation step specifically includes:
[0055] S20: Getting Started
[0056] S21: Determine whether it is the first association? If yes (Y), proceed to S22 and S23, otherwise (N) proceed to S26. The purpose of the first association determination here is to ensure that at least one radar point can update the target motion state;
[0057] S22: Calculate the status flag of the left road boundary object and the positive radar point azimuth of the target;
[0058] S23: Calculate the state flag of the right road boundary object and the negative radar point azimuth of the target;
[0059] S24: Determine whether the distance between the state mark of the left / right road boundary object and the positive / negative radar point is within a predetermined distance? If yes (Y), proceed to S25; otherwise (N), proceed to S26;
[0060] S25: deleting the unassociated radar point (i.e., limiting the target to be associated with the road boundary);
[0061] S26: End.
[0062] In S22 and S23, the distance to the target is approximately calculated using a sine trigonometric function according to the distance and azimuth of the road boundary object.
[0063] Here, another specific example is given. When updating the target, all positions and target objects are compared. Here, there are two thresholds to determine whether the position should be associated. The first threshold is the speed threshold, in this case, whether the position or the target is stationary, this threshold will be easily passed. The second threshold is the radial distance threshold. Since the object is almost close to the road, the radial distance should be handled very carefully, that is, the boundary between the road boundary and the target object needs to be separated. Consider that all stationary objects have two features: the road boundary on the right and the road boundary on the left. At first, the values of both features are false (no), but when they become true (yes), for example, if the road boundary on the right becomes true (yes), their distance difference on the y axis should be calculated. If this distance difference exceeds the preset second threshold (for example, 0.8 meters), this position is discarded. The specific value used as the second threshold here can be adjusted and set according to reference factors such as the road, the road boundary, and the vehicle itself, and the specific value is not limited here.
[0064] As described above, the method for optimizing the association relationship between a target and a road boundary according to the present invention can prevent the position of the road boundary from being erroneously associated with the target, as shown in FIG1(b), which is a schematic diagram showing the association relationship between the target and the road boundary after the present invention is optimized. In FIG1(a) and FIG1(b), the radar point 72 in the two figures is connected by arrows for easy comparison. Compared with FIG1(a), the erroneous association between the radar point 72 and the road boundary is deleted in FIG1(b), that is, the radar point 72 will not be erroneously associated with the real vehicle target. Therefore, according to the present invention, it is possible to effectively prevent roadside static obstacles from being erroneously associated with the road boundary, and it is possible to effectively improve the obstacle detection accuracy, especially the static obstacles.
[0065] Figure 5 It is a structural block diagram of a system for optimizing the association relationship between a target and a road boundary of the present invention.
[0066] like Figure 5 As shown, the system for optimizing the association relationship between the target and the road boundary of the present invention includes:
[0067] A first determination module 100 determines that the target is close to a road boundary based on the number of road boundaries within a predetermined distance around the target;
[0068] A second determination module 200 determines whether the target is close to a left road boundary or a right road boundary; and
[0069] The association restriction module 300 restricts whether the object is associated with the left or right road boundary based on the distance between the object and the left or right road boundary.
[0070] As an example, the first determining module 100 performs the following actions:
[0071] Obtain radar reflection points of the target and the road boundary;
[0072] Determining that the target is a static target;
[0073] Counting the number of radar reflection points of a road boundary within a predetermined distance around the radar reflection point of the target; and
[0074] When the number of radar reflection points of the road boundary within a predetermined distance around the radar reflection point of the target exceeds a preset first threshold, it is determined that the target is close to the road boundary.
[0075] Further, in the first determination module 100, the number of radar reflection points at the road boundary within a predetermined distance around the radar reflection point of the target is counted to obtain a first count value, and when the first count value exceeds the first threshold, it is determined that the target is close to the road boundary.
[0076] Further, in the first determination module 100, the number of radar reflection points of the road boundary existing within a predetermined distance around the radar reflection point of the target on the left and right sides is counted as the first count value to obtain a left count value and a right count value, and when any one of the left count value and the right count value exceeds a first threshold, it is determined that the target is close to the road boundary.
[0077] As a preferred example, when the left side count value exceeds the first threshold value in the first determination module 100, the target is determined to be close to the road boundary on the left side in the second determination module 200; when the right side count value exceeds the first threshold value in the first determination module 100, the target is determined to be close to the road boundary on the right side in the second determination module 200.
[0078] Furthermore, in the association restriction module 300 , when the radial distance between the target and the left or right road boundary is greater than a preset second threshold, the target is restricted from being associated with the left or right road boundary.
[0079] Among them, when it is determined in the second determination module 200 that the target is close to the left road boundary, when the radial distance between the target and the left road boundary is greater than a preset second threshold, the target is restricted to be associated with the left road boundary in the association restriction module 300; when it is determined in the second determination module 200 that the target is close to the right road boundary, when the radial distance between the target and the right road boundary is greater than a preset second threshold, the target is restricted to be associated with the right road boundary in the association restriction module 300.
[0080] The above examples mainly illustrate the method for optimizing the association relationship between the target and the road boundary and the system for optimizing the association relationship between the target and the road boundary of the present invention. Although only some of the specific embodiments of the present invention are described, it should be understood by those skilled in the art that the present invention can be implemented in many other forms without departing from its subject matter and scope. Therefore, the examples and embodiments shown are regarded as illustrative rather than restrictive, and the present invention may cover various modifications and substitutions without departing from the spirit and scope of the present invention as defined in the appended claims.
Claims
1. A method for optimizing the association relationship between a target and a road boundary, characterized in that: include: A first determining step of determining that the target is close to a road boundary based on the number of road boundaries within a predetermined distance around the target; A second determination step is to determine whether the target is close to a left road boundary or a right road boundary; as well as The association restriction step restricts whether the object is associated with the left road boundary or the right road boundary based on the distance between the object and the left road boundary or the right road boundary.
2. The method for optimizing the association relationship between the target and the road boundary according to claim 1, characterized in that: The first determining step comprises: Obtain radar reflection points of the target and the road boundary; Determining that the target is a static target; Counting the number of radar reflection points of a road boundary within a predetermined distance around the radar reflection point of the target; and When the number of radar reflection points of the road boundary within a predetermined distance around the radar reflection point of the target exceeds a preset first threshold, it is determined that the target is close to the road boundary.
3. The method for optimizing the association relationship between the target and the road boundary according to claim 2, characterized in that: The calculating the number of radar reflection points of the road boundary within a predetermined distance around the radar reflection point of the target comprises: Counting the number of radar reflection points at the road boundary that exist within a predetermined distance around the radar reflection point of the target to obtain a first count value; and When the first count value exceeds the first threshold, it is determined that the target is close to a road boundary.
4. The method for optimizing the association relationship between the target and the road boundary as claimed in claim 3, characterized in that: The calculating the number of radar reflection points of the road boundary within a predetermined distance around the radar reflection point of the target comprises: Counting the number of radar reflection points of the road boundary existing within a predetermined distance around the radar reflection point of the target on the left and right sides with the rear axle of the vehicle as the center to obtain a left count value and a right count value as the first count value; and When any one of the left count value and the right count value exceeds the first threshold, it is determined that the target is close to the left road boundary or the right road boundary.
5. The method for optimizing the association relationship between the target and the road boundary according to claim 4, characterized in that: In the case where the left side count value exceeds the first threshold value in the first determining step, determining in the second determining step that the target is a road boundary close to the left side; In the case where the right side count value exceeds the first threshold value in the first determination step, it is determined in the second determination step that the target is a road boundary close to the right side.
6. The method for optimizing the association relationship between the target and the road boundary according to claim 5, characterized in that: In the association restriction step, when the radial distance between the target and the left or right road boundary is greater than a preset second threshold, the target is restricted from being associated with the left or right road boundary.
7. The method for optimizing the association relationship between the target and the road boundary according to claim 6, characterized in that: In the case where it is determined in the second determining step that the target is close to the road boundary on the left, then in the association limiting step, when the radial distance between the target and the road boundary on the left is greater than a preset second threshold, the target is limited to be associated with the road boundary on the left; When it is determined in the second determination step that the target is close to the right road boundary, in the association restriction step, when the radial distance between the target and the right road boundary is greater than a preset second threshold, the target is restricted from being associated with the right road boundary.
8. A system for optimizing the association between a target and a road boundary, characterized in that: include: A first determination module determines that the target is close to a road boundary based on the number of road boundaries within a predetermined distance around the target; A second determination module determines whether the target is close to a left road boundary or a right road boundary; as well as The association restriction module restricts whether the object is associated with the left or right road boundary based on the distance between the object and the left or right road boundary.
9. The system for optimizing the association relationship between the target and the road boundary according to claim 8, characterized in that: The first determination module performs the following actions: Obtain radar reflection points of the target and the road boundary; Determining that the target is a static target; Counting the number of radar reflection points of a road boundary within a predetermined distance around the radar reflection point of the target; and When the number of radar reflection points of the road boundary within a predetermined distance around the radar reflection point of the target exceeds a preset first threshold, it is determined that the target is close to the road boundary.
10. The system for optimizing the association relationship between the target and the road boundary according to claim 9, characterized in that: In the first determination module, the number of radar reflection points at the road boundary within a predetermined distance around the radar reflection point of the target is counted to obtain a first count value, and when the first count value exceeds the first threshold, it is determined that the target is close to the road boundary.
11. The system for optimizing the association relationship between the target and the road boundary according to claim 10, characterized in that: In the first determination module, the number of radar reflection points of the road boundary existing within a predetermined distance around the radar reflection point of the target on the left and right sides is counted with the rear axle of the vehicle as the center to obtain a left count value and a right count value as the first count value, When any one of the left count value and the right count value exceeds the first threshold, it is determined that the object is close to the road boundary.
12. The system for optimizing the association relationship between the target and the road boundary according to claim 11, characterized in that: When the left side count value exceeds the first threshold value in the first determination module, the target is determined to be a road boundary close to the left side in the second determination module; When the right count value exceeds the first threshold value in the first determination module, the second determination module determines that the target is a road boundary close to the right.
13. The system for optimizing the association relationship between the target and the road boundary according to claim 12, characterized in that: In the association restriction module, when the radial distance between the target and the left or right road boundary is greater than a preset second threshold, the target is restricted from being associated with the left or right road boundary.
14. The system for optimizing the association relationship between the target and the road boundary according to claim 13, characterized in that: When it is determined in the second determination module that the target is close to the road boundary on the left, and when the radial distance between the target and the road boundary on the left is greater than a preset second threshold, the target is restricted in the association restriction module to be associated with the road boundary on the left; When the second determination module determines that the target is close to the right road boundary, and when the radial distance between the target and the right road boundary is greater than a preset second threshold, the target is restricted to be associated with the right road boundary in the association restriction module.
15. A computer readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for optimizing the association relationship between a target and a road boundary as described in any one of claims 1 to 7 is implemented.
16. A computer device comprising a storage module, a processor, and a computer program stored in the storage module and executable on the processor, characterized in that: When the processor executes the computer program, the method for optimizing the association relationship between the target and the road boundary according to any one of claims 1 to 7 is implemented.
17. An automatic emergency braking system, characterized in that: A system for optimizing the association relationship between a target and a road boundary comprising the system described in any one of claims 8 to 14.
