Object Recognition Device Dynamic Integration Range
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object recognition systems for vehicles face decreased accuracy due to inappropriate setting of integration determination ranges, which leads to erroneous recognition of objects as either separate or the same, as these ranges are often set along the vehicle's traveling direction rather than the objects' actual trajectories.
Innovation Solution
An object recognition device that estimates the traveling direction of detected objects and adjusts the integration determination distance along this direction, rather than the vehicle's longitudinal direction, to accurately determine whether multiple detected objects belong to the same object.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the integration determination range is increased along the vehicle's traveling direction, then the object detection coverage is improved, but the object recognition accuracy deteriorates due to misalignment with objects' actual trajectories
Solution Approach 1:
The integration determination range is made dynamic by adjusting its direction based on the estimated traveling direction of each object. Instead of using a fixed rectangular range aligned with the vehicle's traveling direction, the system rotates the integration determination range to align with the object's movement trajectory, thereby improving recognition accuracy while maintaining adequate coverage.
Solution Approach 2:
The system changes the orientation parameter of the integration determination range from being fixed along the vehicle's traveling direction to being adjustable according to the object's traveling direction. This parameter change allows the integration range to adapt to different object trajectories, resolving the contradiction between coverage and accuracy.
2Device complexity
If a fixed integration determination range is used, then the processing complexity is reduced, but the object recognition accuracy deteriorates due to inability to adapt to different object trajectories
Solution Approach 1:
The system introduces dynamic adjustment of the integration determination range based on object traveling direction estimation. By making the range adaptive rather than fixed, the system achieves better recognition accuracy without significantly increasing processing complexity, as the adjustment is based on straightforward directional estimation.
Data Source
AI summary
An object recognition device includes: a sensor; a storage device; and a processor configured to detect a first object and a second object around the vehicle, initially set, based on a detection position of the first object, an integration determination distance used to determine that the first object and the second object belong to the same object, estimate a traveling direction of the first object based on the sensor detection information, increase the integration determination distance along the traveling direction, after increasing the integration determination distance along the traveling direction, determine, based on the integration determination distance, whether the first object and the second object belong to the same object, and output the first object and the second object as the same object when it is determined that the first object and the second object belong to the same object.