Fusion detection system, vehicle driver assistance system, and motor vehicle

Abstract

A fusion detection system, a vehicle driver assistance system using the fusion detection system, and a motor vehicle. The fusion detection system comprises: at least one first detector, which gives two object identities ME ID0 and ME ID1 to one actual object detected in a region under detection in one scenario, and obtains a signal value of each of the object identities ME ID0 and ME ID1; at least one second detector, which gives one object identity RD ID0 to the same object detected in the same region under detection, and obtains a signal value of the object identity RD ID0; and a processing unit, which separately compares the signal value of the object identity RD ID0 obtained by the second detector with the signal value of each of the object identities ME ID0 and ME ID1 obtained by the first detector, and uses the object identity ME ID0 or ME ID1 having a smaller difference in the signal value as a correct object identity.

Description

Fusion detection systems, vehicle driver assistance systems and motor vehicles

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of Chinese patent application No. 202410405929.6 filed with the State Intellectual Property Office of China on April 3, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present invention relates to a fusion detection system, a vehicle driving assistance system and a motor vehicle. Background Art

[0004] In recent years, motor vehicles with vehicle driving assistance functions (such as adaptive high beam function, hereinafter referred to as ADB function) have become popular among users. ADB function mostly uses vision-based detectors (such as driving cameras / cameras) to detect the area to be detected in front of the vehicle (i.e., field of view FOV). Existing detectors include commercially available Camera (hereinafter referred to as ME camera).

[0005] However, the ME camera has the following defects: in some application scenarios (for example, in nighttime application scenarios), the ME camera may assign two object identities ME ID0 and ME ID1 to an actual target detected in the area to be detected, although the signal values ​​of the two object identities ME ID0 and ME ID1 are similar to each other.

[0006] While this flaw doesn't have a fatal impact on the decision-making of the vehicle's onboard computer (i.e., the electronic control unit (ECU)), it could still cause some distress to the vehicle's users. Ideally, a detector would assign only one (not two) object identities to a single target detected in the area to be detected. Therefore, improvements to the existing ADB functional module are desired to eliminate or mitigate the aforementioned flaws of ME cameras. Summary of the Invention

[0007] The present invention is made to solve the above technical problems and potential other technical problems.

[0008] According to one aspect of the present invention, a fusion detection system is provided. The fusion detection system includes: at least one first detector, which assigns two object identities ME ID0 and ME ID1 to an actual target detected in a detection area in a scenario, and obtains a signal value for each object identity ME ID0 and ME ID1; at least one second detector, which assigns an object identity RD ID0 to the same target detected in the same detection area, and obtains a signal value for the object identity RD ID0; and a processing unit, which compares the signal value of the object identity RD ID0 obtained by the second detector with the signal value of each object identity ME ID0 and ME ID1 obtained by the first detector, and adopts the object identity ME ID0 or ME ID1 with the smaller difference in signal value from the object identity RD ID0 as the correct object identity, while not adopting the other object identity ME ID0 or ME ID1 with the larger difference in signal value from the object identity RD ID0. By adopting the above technical solution of the present invention, the above-mentioned defects of ME cameras can be eliminated or compensated.

[0009] Specifically, the first detector can be a vision-based detector, and the second detector can be a radar-based detector. More specifically, the first detector can be a driving camera (such as the ME camera mentioned above), and the second detector can be a driving radar. By adopting the above technical solution of the present invention, the above technical problems can be solved at a low cost and with high reliability using existing resources.

[0010] Specifically, if the first detector detects one target in the area to be detected and the second detector detects two targets in the same area, the detection result of the first detector is adopted, and the detection result of the second detector is not adopted. Furthermore, if the first detector assigns only one object identity to a single target detected in the area to be detected, the detection result of the first detector is adopted. By adopting the above-mentioned technical solution of the present invention, it is possible to make accurate judgments about targets in the area to be detected, thereby adopting the correct detection results.

[0011] According to another aspect of the present invention, a vehicle driving assistance system is provided. The vehicle driving assistance system includes the fusion detection system according to the previous aspect. Optionally, the vehicle driving assistance system includes an adaptive high beam system, and the fusion detection system can be included in the adaptive high beam system.

[0012] According to another aspect of the present invention, a motor vehicle is provided. The motor vehicle includes the vehicle driving assistance system according to the previous aspect. By adopting the above technical solution of the present invention, the above technical problem can be solved at a low cost and with high reliability based on existing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] To facilitate understanding of the present invention, the present invention is described in more detail below based on exemplary embodiments and in conjunction with the accompanying drawings. The same or similar reference numerals are used in the accompanying drawings to represent the same or similar components. It should be understood that the drawings are only schematic, and the sizes and proportions of the components in the drawings are not necessarily accurate.

[0014] FIG1 is a schematic diagram of a vehicle driving assistance function in the prior art.

[0015] FIG. 2 is a schematic diagram of a vehicle driving assistance function according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0016] As shown in Figure 1, a conventional vehicle driving assistance system primarily includes a driving camera, an ADB module, a lighting ECU, and a vehicle controller area network (CAN) bus (not shown). The ADB function generally operates as follows: the ME camera, mounted on the vehicle's front end, detects the area to be detected (i.e., the field of view (FOV)) in front of the vehicle, sends the detected objects to the ADB module, and then sends a detection list to the lighting ECU.

[0017] In certain application scenarios (e.g., nighttime applications), the ME camera may assign two object identities, ME ID0 and ME ID1, to a single target detected in the detection area, even though the signal values ​​of the two object identities, ME ID0 and ME ID1, are similar. As a result, the two object identities, ME ID0 and ME ID1, may be displayed in a real-time animation simulation on the vehicle's center console display.

[0018] While this flaw does not have a fatal impact on the decision-making of the on-board computer (i.e., the electronic control unit ECU), it may still cause some confusion to the user of the motor vehicle. The two objects ID0 and ID1 assigned to a target in the field of view (FOV) shown in Figure 1 are exaggerated. In reality, the two objects ID0 and ID1 can be very close to each other or even completely overlap.

[0019] Ideally, the detector assigns only one (not two) object identities to the actual target detected in the detection area. In this way, only one object is displayed in a real-time animated simulation on the vehicle's center console display, thereby matching the actual situation outside the vehicle.

[0020] To this end, the present invention provides a fusion detection system. FIG2 is a schematic diagram of a vehicle driving assistance function according to an exemplary embodiment of the present invention. The vehicle driving assistance system includes an adaptive high beam system, and the fusion detection system can be included in the adaptive high beam system.

[0021] As shown in FIG. 2 , the vehicle driving assistance system according to the exemplary embodiment of the present invention differs from the example in FIG. 1 mainly in that a radar-based detector (eg, a driving radar) is further added.

[0022] The radar can be positioned at the front or corners of the vehicle, depending on the vehicle's specific configuration. Ideally, the radar's field of view (FOV) should be identical to or overlap with that of the existing ME camera. However, in practice, the FOVs of the radar and ME camera may differ. Due to differences in detection principles and FOVs, detection results from the radar and ME camera may differ, requiring further evaluation, as detailed later.

[0023] Similar to the example in FIG. 1 , in certain application scenarios (eg, nighttime application scenarios), the ME camera may assign two object identities ME ID0 and ME ID1 to an actual target detected in the area to be detected.

[0024] At the same time, the radar assigns an object identity (for example, RD ID 0) to the same target detected in the same detection area and obtains a signal value for the object identity RD ID 0. The "signal value" here may include or represent the target's distance, angle, acceleration, confidence level, etc.

[0025] The processing unit (not shown) compares the signal value of the object identity RD ID0 acquired by the radar with the signal value of each of the object identities ME ID0 and ME ID1 acquired by the ME camera, and adopts the one of the object identities ME ID0 and ME ID1 whose signal value has a smaller difference with the object identity RD ID0 as the correct object identity, while not adopting the other one of the object identities ME ID0 and ME ID1 whose signal value has a larger difference with the object identity RD ID0.

[0026] Specifically, as shown in FIG2 , the difference between object identity RD ID0 and object identity ME ID0 is small (visually shown as the distance between object identity RD ID0 and object identity ME ID0 is relatively close), while the difference between object identity RD ID0 and object identity ID1 is relatively large (visually shown as the distance between object identity RD ID0 and object identity ME ID1 is relatively far). Therefore, in this case, object identity ME ID0 is adopted as the correct object identity, while the other object identity ME ID1 is not adopted. It should be understood that the various object identities in the field of view FOV shown in FIG2 are exaggerated, and in reality, these object identities may be very close to each other or even overlap.

[0027] In addition, if the ME camera detects one target in the detection area, and the radar detects two targets (ID0 and ID2) in the same detection area, the ME camera's detection result is used instead of the radar's. This may be due to the difference in detection range between the ME camera and the radar.

[0028] In addition, in the case where the ME camera assigns only one object identity to actually one target detected in the area to be detected, the detection result of the ME camera is adopted because this is exactly the desired ideal detection result.

[0029] In summary, the judgment logic of the vehicle driving assistance function of the present invention can be listed as follows:

[0030] IF only one target is detected in the field of view FOV of the ME camera and two object identities ME ID0 and ME ID1 are found

[0031] ELSE IF Only one target is detected by the radar

[0032] THEN 1. Compare the signal value of object identity RD ID0 with the signal values ​​of two object identities ME ID0 and ME ID1 respectively

[0033] 2. Difference ΔME ID0 = (Signal value of ME ID0 - Signal value of RD ID0) / Signal value of ME ID0

[0034] 3. Difference ΔME ID1 = (Signal value of ME ID1 - Signal value of RD ID0) / Signal value of ME ID1

[0035] 4. Output the smaller of the difference ΔME ID0 and the difference ΔME ID1 to the vehicle CAN

[0036] ELSE IF Two targets are detected by radar

[0037] THEN outputs the signal values ​​of ME ID0 and ME ID1

[0038] ELSE Only one target is detected in the field of view FOV of the ME camera and one object identity ME ID0 is found

[0039] THEN output the signal value of ME ID0 to the vehicle CAN

[0040] Although the technical objectives, technical solutions, and technical effects of the present invention have been described in detail above with reference to specific embodiments, it should be understood that the above embodiments are merely illustrative and not restrictive. Within the essential principles of the present invention, any modifications, equivalent substitutions, and improvements made by those skilled in the art are intended to be included within the scope of protection of the present invention.

Claims

1. A fusion detection system, characterized in that: include: at least one first detector, which assigns two object identities ME ID0 and ME ID1 to an actual target detected in an area to be detected in a scenario, and obtains a signal value of each object identity ME ID0 and ME ID1; at least one second detector, which assigns an object identity RD ID0 to the same target detected in the same area to be detected, and obtains a signal value of the object identity RD ID0; as well as a processing unit, the processing unit comparing the signal value of the object identity RD ID0 acquired by the second detector with the signal values ​​of each object identity ME ID0 and ME ID1 acquired by the first detector, and adopting the one of the object identities ME ID0 and ME ID1 having a smaller difference in signal value with the object identity RD ID0 as the correct object identity, and not adopting the other one of the object identities ME ID0 and ME ID1 having a larger difference in signal value with the object identity RD ID0.

2. The fusion detection system according to claim 1, wherein: The first detector is a vision-based detector and the second detector is a radar-based detector.

3. The fusion detection system according to claim 1, wherein: The first detector is a driving camera, and the second detector is a driving radar.

4. The fusion detection system according to claim 1, wherein: In the case where the first detector detects one target in the area to be detected and the second detector detects two targets in the same area to be detected, the detection result of the first detector is adopted, and the detection result of the second detector is not adopted.

5. The fusion detection system according to claim 1, wherein: In the case that the first detector assigns only one object identity to an actual target detected in the area to be detected, the detection result of the first detector is adopted.

6. A vehicle driving assistance system, characterized in that: The vehicle driver assistance system comprises a fusion detection system according to any one of the preceding claims.

7. The vehicle driving assistance system according to claim 6, wherein: The vehicle driving assistance system includes an adaptive high beam system, and the fusion detection system is included in the adaptive high beam system.

8. A motor vehicle, characterized in that: The motor vehicle comprises a vehicle driving assistance system according to claim 6 or 7.

Images