道路抛洒物检测方法、装置及可读存储介质
By combining semantic segmentation and classification models, the accuracy and efficiency issues of road debris detection are solved, achieving efficient identification of road debris and reducing system load.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CENNAVI TECH
- Filing Date
- 2023-06-07
- Publication Date
- 2026-07-17
AI Technical Summary
In existing technologies, the detection of road debris relies on manual patrols, which is costly and cannot be processed in a timely manner. Furthermore, existing algorithm models have low accuracy and are difficult to effectively identify the type and location of road debris.
This method combines semantic segmentation and classification models. By acquiring the image to be identified, the driving area and the true outline of the object to be identified are determined. It then determines whether the object is located within the driving area and uses a pre-set classification model to determine the probability score of the object being a spill, thereby improving the recognition accuracy.
It enables accurate identification of road debris, reduces system load, improves detection accuracy, and reduces the need for manual inspections.
Smart Images
Figure CN116758512B_ABST