Rear Cross-Traffic Warning Using Camera-Radar Static Object Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Radar-based rear cross-traffic collision warning systems frequently generate false warnings due to the inability to differentiate between static and moving objects, especially when a static object is present in the vehicle's direction of travel.
Innovation Solution
A vehicle collision avoidance system utilizing a camera to recognize static objects and determine a free space, combined with radar to detect moving objects, determines whether to warn of a collision based on the relative positions and distances of these objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If radar-based detection is used to detect objects at the rear side, then moving objects can be detected, but false warnings occur when static objects are present in the direction of travel
Solution Approach 1:
The patent introduces a camera as an intermediary device to capture images of the rear environment. The captured images are processed to identify static objects (such as pillars, walls, or parked vehicles) that are not actually moving threats. This intermediary visual verification mechanism filters out false positives from radar detection, resolving the contradiction between detecting moving objects and avoiding false warnings from static objects.
2Measurement precision
If radar sensor is used for rear cross-traffic collision warning, then approaching vehicles can be detected, but the system cannot differentiate between static and moving objects
Solution Approach 1:
The patent merges radar detection capabilities with camera imaging capabilities into a fused detection system. The radar provides distance and velocity information while the camera provides visual identification and motion context. By combining these complementary sensing modalities, the system achieves both precise object detection and accurate determination of motion state, resolving the contradiction between detection accuracy and motion information loss.
3Reliability
If the system warns of collision based on radar detection alone, then safety is improved, but unnecessary braking occurs due to false warnings
Solution Approach 1:
The system implements a feedback mechanism where camera images are continuously monitored to verify the nature of detected objects. When radar detects a potential threat, the camera provides feedback to confirm whether the object is actually moving toward the vehicle. This feedback loop prevents unnecessary braking actions by validating radar alerts against visual evidence, thus maintaining both safety and driving smoothness.
Data Source
AI summary
A vehicle collision avoidance device and a method therefor are provided. The vehicle collision avoidance device includes: a camera that captures a rear side image of a vehicle; a radar that recognizes an object present at a rear side of the vehicle; and at least one processor electrically connected to the camera and the radar. The at least one processor: determines whether there is a static object at the rear side of the vehicle; determines a position of the static object present at the rear side of the vehicle; determines an empty space present in the direction of travel of the vehicle, based on the rear side image obtained by the camera and the position of the static object; and determines whether to warn of a collision, based on a moving object and the static object, when the moving object is detected.