Autonomous Parking Stop Position Determination Using Multi-Camera Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In autonomous parking systems, sudden changes in target parking position occur due to differences in performance and position of multiple shooting units, leading to unstable vehicle behavior.
Innovation Solution
An image processing device comprising a first generation unit, a second generation unit, and a determination unit that generates first and second target parking frame information from side and rear cameras respectively, and determines a stop position by integrating both to prevent sudden changes in the target parking position.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple shooting units are used to detect parking frames, then detection coverage and reliability are improved, but sudden changes in target parking position occur due to differences in shooting unit performance and positions
Solution Approach 1:
The patent combines detection results from multiple shooting units (side camera and rear camera) by integrating their respective target parking frame information. The determination unit synthesizes first target parking frame information from the side camera and second target parking frame information from the rear camera to generate third target parking frame information, thereby preventing sudden changes in target parking position while maintaining detection reliability.
2Adaptability or versatility
If shooting units are switched based on parking environment changes, then adaptability to different distances is improved, but target parking position changes suddenly due to performance differences between units
Solution Approach 1:
The patent implements a dynamic switching mechanism where the determination unit selectively uses detection results from the side camera or rear camera based on the detected parking environment and distance. This dynamic approach allows the system to adapt to different parking scenarios while maintaining target parking position continuity through proper integration of detection results from different shooting units.
3Device complexity
If autonomous parking control uses a single shooting unit, then system complexity is reduced, but detection accuracy and coverage are limited
Solution Approach 1:
The patent makes the image processing device multi-functional by enabling it to process detection results from multiple types of shooting units (side camera and rear camera) for different parking scenarios. The determination unit universally handles integration of detection results from either camera, allowing the system to maintain high detection accuracy and coverage while managing complexity through a unified processing approach.
Data Source
AI summary
An image processing device includes: a first generation unit which generates first target parking frame information based on an image taken by a first shooting unit that is provided at a side position of a vehicle; a second generation unit which generates second target parking frame information based on an image taken by a second shooting unit that is provided at a front or rear position of the vehicle; and a determination unit which determines a stop position in a parking frame by generating third target parking frame information to serve as a stop target based on the first target parking frame information generated by the first generation unit and the second target parking frame information generated by the second generation unit.