Research On Intelligent Towing Induction Technology Of Road Rescue Equipment Based On Vision

With the continuous growth of social economy,the number of motor vehicles presents explosive growth,road traffic accidents occur frequently,and the road rescue industry gradually receives social attention.As the most frequently used basic operation in the rescue process,towing is difficult to operate.However,there is still a lack of research on towing induction technology of road rescue equipment at home and abroad.The towing operation has low intelligence.In the process of towing rescue,towing operation is often time-consuming and inefficient,resulting in traffic congestion and even secondary accidentsIn view of the problems existing in the current towing induction technology in the field of road rescue,this paper conducts a comprehensive study on the towing induction technology under two towing conditions,namely positive and lateral directions.The thesis has completed the following major works:(1)Based on the analysis of the difficult problems of road rescue and towing work and the research status of related technologies,the overall scheme of road rescue equipment intelligent towing guidance system based on vision is designed.(2)According to the structure and working characteristics of road rescue equipment,an intelligent towing guidance system of road rescue equipment based on vision is constructed,and the hardware architecture and software platform of the system are designed and built(3)Combined with the positive azimuth towing operation scene,a positive azimuth towing induction method based on license plate recognition was proposed,including the license plate recognition algorithm based on license plate shape and corner feature and the lateral deviation measurement method of positive azimuth towing.(4)Combined with the scene of lateral azimuth towing operation,a lateral azimuth towing induction method based on wheel recognition was proposed,including the wheel recognition and tracking algorithm based on kalman filter and the lateral deviation measurement method of lateral azimuth towing.(5)Aiming at the shortcomings of traditional machine vision methods such as manual feature extraction and limited adaptability to the environment,the license plate recognition technology based on deep learning was studied.A tiny-yolo based license plate recognition method was improved and the effect of network model was preliminarily verified by simulation experiment.(6)In order to verify the function and reliability of intelligent towing induction system for road rescue equipment,a real vehicle comprehensive test was carried out in the test field,and the test results were analyzed.