基于图像识别的路面健康状态检测方法
By using multi-frame image processing and structural perturbation constraint models, the problem of difficulty in identifying minor road surface defects in low-speed, heavy-load scenarios in existing technologies has been solved, achieving high-precision detection and instability risk assessment of early cracks and oil film anomalies.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SICHUAN KANGJISHENG TECHNOLOGY CO LTD
- Filing Date
- 2026-05-09
- Publication Date
- 2026-07-17
AI Technical Summary
Existing manual inspections and detection methods based on thresholds for large cracks and potholes are insufficient to reliably identify minor, high-risk pavement defects in low-speed, heavy-load scenarios such as downhill ramps at bridge-tunnel junctions and braking zones before toll stations, under complex lighting conditions, water reflection, and road marking interference. These defects include early network cracking, shot blasting spalling, and localized oil seepage, making it impossible to determine the pavement health status in a timely and accurate manner.
By acquiring multiple frames of images of the same road surface area of the target road segment during continuous vehicle travel, the inter-frame projection correlation is established, reflective information is separated, a structural disturbance constraint model is constructed, and markings, manhole cover edges, etc. are identified. Subpixel crack extraction and oil film anomaly identification are performed. Combined with the co-position retention and spread analysis of multiple frames, the damage type and instability risk are determined.
It effectively reduces the false detection rate of defects under complex lighting and rainy night reflective conditions, improves the accuracy of identifying millimeter-level early cracks, localized aggregate exposure, and abnormal oil film, and achieves an improvement from defect identification to instability risk assessment. It is suitable for accurate detection in low-speed, heavy-load, and slippery, easily unstable scenarios.
Smart Images
Figure CN122150271B_ABST