Research on Intelligent Traffic Inspection Applications Based on Unmanned Aerial Vehicle Imaging
TL;DR
Experimental results indicate that the proposed automated inspection solution can effectively identify and track different types of vehicles in port traffic management scenarios, validating its accuracy and reliability in practical applications and providing a new solution for intelligent traffic inspection.
Abstract
: This paper explores the applications of intelligent traffic inspection based on unmanned aerial vehicle (UAV) imaging. Addressing the issues of low efficiency, high safety risks, and limited coverage in traditional traffic inspection methods, it proposes an automated inspection solution that integrates UAV technology with image recognition technology. The study employs a DJI M300 RTK UAV equipped with visual sensors for data collection, utilizing the YOLO 11 algorithm to achieve efficient and precise recognition of vehicles and traffic anomalies. Imaging optimization of the visual sensors has been carried out to ensure image quality under varying weather and lighting conditions. Compared to its predecessors, the YOLO 11 algorithm demonstrates significant improvements in inference speed, recognition accuracy, occlusion tolerance, and model lightweightness, making it particularly suitable for small object recognition tasks in traffic inspection. Through the design of a standardized UAV traffic inspection process, full-chain automated processing is realized, covering video collection, quality verification, object detection and tracking, to result output and data archiving. Experimental results indicate that the proposed solution can effectively identify and track different types of vehicles in port traffic management scenarios, validating its accuracy and reliability in practical applications and providing a new solution for intelligent traffic inspection.