Font Size: a A A

Research On Vehicle Blind Spot Monitoring And Warning System Based On DSP

Posted on:2017-11-23Degree:MasterType:Thesis
Country:ChinaCandidate:S LiFull Text:PDF
GTID:2492305348495594Subject:Vehicle Engineering
Abstract/Summary:PDF Full Text Request
With the development of the automotive industry,the number of vehicles are increasing rapidly,and so is the number of the traffic accidents.As a result,people have attached more importance to vehicle safety systems(active safety systems and passive safety systems).But the development of the traditional passive safety systems has met its bottleneck,while the active safety systems have been thriving.Due to the limitation of the rearview mirrors,there exists a blind zone where the drivers cannot see this area.And therefore,a lot of accidents happened because of this.The blind spot monitoring(BSM)system has been designed to fix this.In order to help the driver to detect the potential danger of the blind zone and increase driving safety,this paper has studied a vehicle blind spot monitoring and warning system based on DSP.This paper analyzes the feasibility and necessity of the BSM system,explores the present situation of domestic and foreign researches and finds their advantages and shortages.Based on the needs that the society demands for the BSM system,this paper combines the software design,hardware design,recognition and tracking algorithm and warning strategy to BSM warning system using a single vision camera,proposes a feasible solution for the BSM system.The algorithm of this paper is based on the Harris corner detection method,the Scale Invariant Feature Transform(SIFT)and the Optical Flow algorithm.Because the Harris corner is the relatively stable features and the Harris and SIFT features are both rotation invariant,this algorithm uses Harris algorithm to detect the feature point and obtain the feature vector by the SIFT direction distribution and descriptor algorithm.Then does the feature matching during the image sequence.This paper also has studied the optical flow based moving vehicle tracking method.Set the region of interest(ROI)to reduce the computational complexity to improve the efficiency of the algorithm.This paper has also designed the embedded version of the system based on a digital signal processor,and has worked on the optimization followed by the instruction and datasheet of the core DSP.After some experiments on the real road,the system meets the requirements of the system functional design.
Keywords/Search Tags:feature extraction, sparse optical flow, blind spot monitoring (BSM), ADSP, machine vision
PDF Full Text Request
Related items