基于毫米波雷达与视觉融合的超限车辆车道级识别方法
By synchronously triggering and fusing the cascaded millimeter-wave radar array with the camera, the problem of insufficient lane-level positioning accuracy was solved, enabling real-time lane-level recognition and overload vehicle detection on roadside edge computing devices, and adapting to lane determination for different road shapes.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANDONG BOAN INTELLIGENT TECH CO LTD
- Filing Date
- 2026-04-21
- Publication Date
- 2026-07-17
AI Technical Summary
Existing radar and vision fusion solutions are insufficient in terms of lane-level positioning accuracy, especially prone to misjudgment on curved or gradually changing road sections, and have high computational overhead, making it difficult to meet real-time requirements.
The system employs a cascaded millimeter-wave radar array and a camera for synchronous triggering. Through channel calibration, frequency domain transformation, and constant false alarm rate detection, it extracts the azimuth angle and performs spatial clustering. Combined with visual observations, it performs frame-by-frame pairing and association. Using a multinomial fitting and voting mechanism, it determines lane affiliation and outputs lane-level identification results for overloaded vehicles.
It significantly improves the angular resolution of radar, reduces lane-level misjudgment rate, enables real-time processing on roadside edge computing devices, adapts to lane-level recognition of different road shapes, and provides complete vehicle outline size estimation and over-limit judgment.
Smart Images
Figure CN122116652B_ABST