Highway foreign matter identification and hierarchical early warning method and system based on multi-modal data enhancement and adaptive fusion

CN121236717BActive Publication Date: 2026-06-09安徽交控工程集团有限公司

Patent Information

Authority / Receiving Office: CN · China
Patent Type: Patents(China)
Current Assignee / Owner: 安徽交控工程集团有限公司
Filing Date: 2025-10-14
Publication Date: 2026-06-09

AI Technical Summary

Technical Problem

Existing single-modal sensing technologies are susceptible to interference from changes in lighting and weather conditions in highway foreign object detection, and it is difficult to achieve a dynamic balance between recognition accuracy and computational efficiency, thus failing to meet the requirements of real-time performance and resource consumption.

Method used

A multimodal data augmentation and adaptive fusion method is adopted. By simultaneously collecting LiDAR point cloud data and camera image data, preprocessing and spatiotemporal alignment are performed, multi-scale geometric and visual features are extracted in parallel, the optimal model is selected by a dynamic model arbiter, and heterogeneous feature fusion is performed through feature gating mechanism to achieve foreign object identification and hierarchical early warning.

Benefits of technology

It achieves high-precision, robust, and adaptive foreign object identification and early warning in complex and ever-changing highway scenarios, meets the real-time processing requirements in resource-constrained environments, and reduces false alarm and false negative rates.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure CN121236717B_ABST

Patent Text Reader

Abstract

The present application relates to a kind of multi-modal data enhancement and adaptive fusion expressway foreign matter identification and grading early warning method and system.The method includes: laser radar point cloud and camera image data are synchronously collected and preprocessed and space-time alignment are carried out;Parallel extraction point cloud multiscale geometric feature and image multiscale visual feature;According to real-time scene and system state, select optimal model combination from pre-configured model pool by dynamic model adjudicator;Foreign matter identification is completed by using feature gate mechanism to fuse heterogeneous features;Based on the identification result, risk classification is carried out using a multi-factor weighted decision model and an early warning signal is triggered.The present application significantly improves the accuracy, robustness and real-time performance of expressway foreign matter identification through multi-modal data collaborative processing and adaptive fusion mechanism.

Need to check novelty before this filing date? Find Prior Art