Font Size: a A A

Research On Traffic Sign Detection System Based On Auxiliary Driving

Posted on:2020-04-09Degree:MasterType:Thesis
Country:ChinaCandidate:J LiFull Text:PDF
GTID:2392330602964324Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of China's automobile industry,the traditional operation management model cannot effectively solve the dilemma faced by road traffic,while the emergence and development of Advanced Driver Assistance Systems(ADAS)in the artificial intelligence industry provides a new solution to solve the road traffic dilemma.The perfect landing of all functions in the auxiliary driving system is the prerequisite for the realization of automatic driving.As an important part of ADAS,the traffic sign detection system plays an important role in improving driving safety performance and solving traffic operation efficiency.There are many kinds of traffic signs,but injunctions have a prohibition or restriction effect on the behavior of road participants,and they have a more significant impact on driving safety and road operation efficiency.This paper makes an in-depth study of the detection and recognition methods of circular injunctions in natural scenes by using on-board visual system,including software system design and hardware system design.Main research contents are as follows:(1)the software part:In view of the research status at home and abroad,this paper proposes a method based on MSER to extract candidate regions from grayscale images to achieve rough selection of target regions.Then the target region is extracted according to the circular geometric features to realize the accurate detection and segmentation of the prohibition mark in the image.Finally,to improve the real-time performance of the detection algorithm,a region target tracking detection algorithm is proposed.Sign recognition process,the first use of Gradient direction Histogram(the Histogram of Oriented Gradient,HOG)complete ban on circular signs feature extraction,then multidimensional feature Vector to the classifier of Support Vector Machine(Support Vector Machine,SVM),complete the classifier training,ban sign is realized by using the trained classifier recognition classification.(2)the hardware part:The traffic sign detection system is installed on the front windshield of the car in the form of vehicle-mounted vehicles.The main work of this paper is to design the visual detection system.Through the detection system fixed on the vehicle,image collection,sign detection and recognition are completed successively.The principle of fixed device design is to prevent blurred images from being collected in the process of vehicle driving due to turbulence and vibration,realize anti-shake and de-shake on the mechanical structure,and simplify the next ph ase of the ban sign detection and recognition algorithm.
Keywords/Search Tags:Assisted driving, Prohibition traffic sign, Circle detection, Recognition, SVM
PDF Full Text Request
Related items