Vehicle Control Device Using Estimation Line for Landmark Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle control devices using wide-angle radar systems, such as millimeter wave radar and LIDAR, often incorrectly recognize roadside objects installed at equal intervals as vehicles traveling side-by-side, leading to misunderstandings about lane interruptions.
Innovation Solution
A vehicle control device incorporating a wide-angle radar, map database, estimation line setting, and object estimation processes to differentiate between roadside objects and actual vehicles by determining positional relationships and relative speeds, using estimation lines that tilt with distance from the vehicle and determination speed areas to accurately assess landmarks for travel assist control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If wide-angle radar is used to detect landmarks around the own vehicle, then the detection coverage is improved, but roadside objects installed at equal intervals are incorrectly recognized as vehicles traveling side-by-side
Solution Approach 1:
The patent introduces an estimation line as an intermediary reference based on borderline information from map data. This estimation line serves as a mediator to distinguish between actual vehicles and roadside objects by comparing the relative positional relationships of detected landmarks against the expected geometric pattern defined by the estimation line extending from the borderline
Solution Approach 2:
The patent transitions from two-dimensional radar detection coordinates to a three-dimensional spatial reasoning framework by incorporating map data (borderline information) to create an estimation line. This adds a geometric constraint dimension that enables discrimination between objects at equal intervals (roadside objects) and actual vehicles based on their positional relationships relative to the road boundary
2Reliability
If the detection sensitivity is increased to detect all potential vehicles, then more vehicles are detected, but false detection of roadside objects increases
Solution Approach 1:
The patent implements a feedback mechanism where the estimated position of the borderline (from map data) and the calculated estimation line provide a reference framework. The actual detected landmark positions are continuously compared against this reference, and the recognition result is fed back to confirm or reject vehicle detection, thereby reducing false detections while maintaining high sensitivity
Data Source
AI summary
It is estimated whether or not each of landmarks is an object to be noted for travel assist control by using an estimation line EL set in the reference frame. The estimation line EL is set to be substantially parallel to a borderline BL. The borderline BL is a line a line which separates a drivable area and an undrivable area of a vehicle. It is estimated that the landmark located on a center line CL side is the object to be noted for travel assist control. It is estimated that the landmark located on a borderline BL side is not the object to be noted for travel assist control.