Stationary Object Detection Grouping for Sparse Sensor Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Techniques using spatially sparse sensors fail to accurately recognize the environment around a vehicle, mistakenly identifying non-travelable areas as travelable due to sparse detection points.
Innovation Solution
An object recognition apparatus that extracts stationary object-detection points from multiple detection timings and groups them to accurately identify stationary objects, using spatial density to distinguish between true and false detections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If spatially sparse sensors are used for obstacle detection, then device complexity is reduced, but measurement precision deteriorates causing inaccurate environment recognition
Solution Approach 1:
The system performs preliminary detection using sparse sensors to identify candidate stationary objects, then pre-processes this information by extracting stationary object-detection points and grouping them before final environment recognition. This preliminary grouping action enables accurate recognition despite sparse sensor input.
2Measurement precision
If sensors are densely distributed to improve detection accuracy, then measurement precision improves, but device complexity increases
Solution Approach 1:
The detection process is segmented into distinct functional units: extraction unit for identifying stationary object-detection points, grouping unit for clustering detection points belonging to the same object, and recognition unit for final environment mapping. This segmentation allows sparse sensors to achieve high precision through sophisticated processing rather than dense physical distribution.
3Measurement precision
If multiple detection points are collected at multiple timings to improve recognition accuracy, then measurement precision improves, but loss of time increases
Solution Approach 1:
The system performs detection at multiple periodic timings, collecting detection points in cycles. At each period, the extraction unit identifies stationary object-detection points and the grouping unit processes them. This periodic operation balances thorough detection with time efficiency, allowing the system to achieve accurate recognition without continuous monitoring.
Data Source
AI summary
An extraction unit (101) extracts as a stationary object-detection point, a detection point on a stationary object among a plurality of detection points around a vehicle (200), the plurality of detection points being detected by an outside-detection sensor (501) at a plurality of detection timings. A grouping unit (105) groups two or more stationary object-detection points deduced as detection points on a same stationary object, among a plurality of stationary object-detection points extracted by the extraction unit (101) at the plurality of detection timings.