Vehicle Object Detection Fusion for Crowded Target Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection devices face reduced accuracy in distinguishing closely spaced objects due to instability in target information from sensors with low separability, leading to destabilized fusion of information from sensors with different resolutions.
Innovation Solution
An object detection device that includes a crowd determination unit to identify crowd targets based on a reference target, a region setting unit to define a search region encompassing these targets, and an object determination unit to fuse target information within this region, ensuring accurate object detection by prioritizing sensors with higher resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If target information from sensors with low separability is used for object detection, then the quantity of detected targets increases, but the stability of target information and fusion accuracy deteriorates
Solution Approach 1:
The patent divides the detection space into multiple search regions based on detection results from sensors with high separability. By segmenting the search space, the system can process targets in different regions using appropriate fusion strategies, maintaining stability while detecting multiple targets including those from sensors with lower separability.
Solution Approach 2:
The patent applies different information fusion strategies to different search regions based on local characteristics. Regions with closely spaced targets use prioritized fusion based on sensor resolution, while other regions can use standard fusion, thereby maintaining overall system reliability while increasing total target detection quantity.
2Adaptability or versatility
If target information from sensors with different resolutions is fused, then the comprehensiveness of object detection improves, but the accuracy of object determination deteriorates due to destabilized fusion
Solution Approach 1:
The patent dynamically adjusts the information fusion strategy based on the spatial distribution of targets and sensor characteristics. When targets are closely spaced, the system dynamically prioritizes information from sensors with higher resolution for those specific regions, while still incorporating data from sensors with lower resolution for other areas, thus maintaining both comprehensiveness and accuracy.
Solution Approach 2:
The patent changes the fusion parameters (prioritization weights) based on target density and spatial relationships. In regions where targets are closely spaced, the system increases the weight of high-resolution sensor data and decreases the weight of low-resolution sensor data, thereby maintaining determination accuracy while still achieving comprehensive detection through multi-sensor fusion.
3Device complexity
If a uniform search region is used for all targets, then the simplicity of processing is maintained, but the accuracy of object detection in crowded regions deteriorates
Solution Approach 1:
The patent segments the uniform search region into multiple sub-regions based on target density and spatial distribution. This segmentation allows the system to apply different fusion strategies to different regions, improving detection accuracy in crowded areas while maintaining relatively simple processing through automated region classification.
Data Source
AI summary
An object detection device detects objects around an own vehicle by fusing a plurality of pieces of target information obtained by detecting the objects as targets using a plurality of detection sensors with different detection accuracies, and includes a crowd determination unit that determines that a target other than a reference target detected in a predetermined crowd region based on the reference target selected from the targets is a crowd target; a region setting unit that sets a search region so as to include a crowd determination region in which the reference target and the crowd target are detected; and an object determination unit that determines an object indicated by a target in the search region by fusing the target information for the search region.