Radar Target Recognition for Single-Sensor Roadside Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing roadside object recognition devices require both radar and camera sensors to accurately identify roadside objects, which can be costly and inefficient.
Innovation Solution
A target recognition device that utilizes a microcomputer with sensors like a radar sensor and vehicle speed/yaw rate sensors to estimate a vehicle's future path, detect targets, determine roadside objects based on specific criteria, calculate lateral coordinates, estimate road boundaries, and increase the likelihood of targets being roadside objects based on distance thresholds without relying on both radar and camera sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If both radar and camera sensors are used to accurately identify roadside objects, then recognition accuracy is improved, but device complexity and cost increase
Solution Approach 1:
The patent makes the radar sensor perform multiple functions: both detecting targets and estimating road boundaries. By calculating lateral coordinates of radar reflection points and comparing them with vehicle traveling paths, the system derives road boundary information from the same sensor used for target detection, eliminating the need for separate camera hardware while maintaining recognition accuracy
Solution Approach 2:
The patent combines target detection and road boundary estimation into a single integrated process using radar data. The reflection point group from radar is simultaneously used for both identifying roadside objects and determining road boundaries through lateral coordinate calculation and traveling path comparison, merging functions that traditionally required separate sensors
2Productivity
If radar reflection points are used to identify roadside objects, then detection speed is improved, but measurement precision deteriorates due to false reflection points
Solution Approach 1:
The patent extracts and removes false reflection points from the radar reflection point group by comparing lateral coordinates with the estimated traveling path. Reflection points that do not align with the road boundary derived from the traveling path are identified and excluded, separating true roadside object reflections from false ones while maintaining fast detection speed
Solution Approach 2:
The system uses the estimated traveling path as feedback to validate radar reflection points. By continuously comparing reflection point lateral coordinates against the predicted vehicle path and road boundary, the system refines its identification of true roadside objects, using the traveling path estimation as a feedback mechanism to filter false detections
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The device effectively recognizes roadside objects and accurately estimates road boundaries, distinguishing between fixed objects and moving vehicles, even when only using a single type of sensor, thereby enhancing recognition accuracy and reducing hardware requirements.
Implementation Method 1
a target detection unit configured to detect a target present ahead of the vehicle by using a sensor
Implementation Method 2
acquires a reflection point group by using the radar
Data Source
AI summary
A target recognition device includes a roadside object determination unit, a coordinate calculation unit, a road boundary estimation unit, a distance calculation unit, and a likelihood increasing unit. The roadside object determination unit determines whether a target has a feature of a roadside object. The coordinate calculation unit calculates a coordinate in a lateral direction of the target. The road boundary estimation unit estimates a road boundary. The distance calculation unit calculates a distance from the target to the road boundary. The likelihood increasing unit increases a likelihood that the target is a roadside object, on condition that the distance is a preset threshold value or less.