Stereo Camera FOV Adjustment for Intersection Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional driving assist systems using stereo cameras have a limited field of view of about 45 degrees, which can miss transversely moving vehicles at intersections, reducing their effectiveness in detecting potential collision hazards.
Innovation Solution
A driving assist system that includes a determiner to identify intersections using camera information from a stereo camera and an adjuster to adjust the field of view of the cameras, allowing for a wider detection area by reducing the overlap between the first and second fields of view, thereby enhancing the system's ability to detect vehicles in right or left turn lanes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the field of view of stereo cameras is maintained at the conventional overlap configuration, then the depth perception and stereo matching accuracy are preserved, but the detection coverage at intersections is limited and cannot detect transversely moving vehicles
Solution Approach 1:
The patent implements dynamic adjustment of camera field of view based on detected road geometry. When an intersection is detected, the system adjusts the FOV to reduce overlap and expand transverse coverage. This dynamic reconfiguration allows the system to adapt its detection pattern to the environmental context, switching between depth-optimized and coverage-optimized modes as needed.
Solution Approach 2:
The system changes the field of view parameters (angular coverage and overlap degree) based on the detected road situation. By modifying these optical parameters dynamically, the system can expand its detection area at intersections while maintaining adequate stereo matching performance through controlled overlap reduction rather than complete elimination.
2Area of stationary object
If the field of view overlap between first and second cameras is reduced to expand detection area, then the coverage at intersections is improved, but the stereo matching precision may be degraded
Solution Approach 1:
The system performs preliminary detection of road geometry and intersection characteristics before finalizing the detection pattern. By identifying intersections in advance, the system can proactively adjust its field of view configuration to optimize for transverse vehicle detection, rather than reacting after missing detections occur.
Solution Approach 2:
The system uses feedback from road geometry detection and initial object detection to dynamically adjust camera field of view parameters. This closed-loop control allows the system to find the optimal balance between coverage area and detection precision based on real-time environmental conditions, adjusting the degree of FOV overlap according to the specific situational requirements.
Data Source
AI summary
The present disclosure provides a driving assist system including: a determiner configured to determine whether a current road is an intersection on the basis of camera information obtained from a stereo camera for monitoring the front area of a vehicle; and an adjuster configured to adjust the direction of a first field of view (FOV) or a second field of view such that the overlap area of the first field of view and the second field of view that are the fields of view of a first camera and a second camera included in the stereo camera is reduced, when the current road is determined as an intersection.