Font Size: a A A

Design And Implementation Of Vehicle Panorama Monitoring Terminal Based On ARM And Android Platform

Posted on:2017-05-31Degree:MasterType:Thesis
Country:ChinaCandidate:L MengFull Text:PDF
GTID:2132330503473361Subject:Engineering
Abstract/Summary:PDF Full Text Request
Due to economic development and the promotion of national policy,there is a great increase in the number of cars,especially private cars,which also lead to the uneven drivers,frequent traffic accidents based on various reasons.The problem of traffic safety is turning more and more prominent.On the other hand,social touch porcelain and other uncivilized phenomenon increases,causing the vehicle monitoring system arising.One-way vehicle monitoring system can provide evidence for a variety of traffic accidents or disputes,but it is only able to be used after the accident instead of avoid an accident.In addition to providing evidence, the panoramic traffic monitoring system can eliminate the blind spot of the driver’s sight, and avoid the occurrence of accidents in advance.This system is based on the Android platform,FreescaleIMX6 processor,TW6865 video capture module and ARM cortex-a9 architecture,with the help of the super wide-angle fish-eye lens. The fish-eye lens can achieve wide sight,but it needs corrected due to the serious image deformation caused by the inherent distortion of the fish-eye lens itself.The geometric projection model of the fish-eye lens is divided into four types, namely, orthogonal projection, isometric projection,solid angle projection and stereo projection.The most widely used is the isometric projection model, but most of correction researches are based on the orthogonal projection model.There are many methods of the correction for fish-eye image,common with the correction algorithms based on lens calibration,polynomial coordinate transformation and projection model.This paper introduces several correction model,and put forward the correction model with the help of the isometric projection of spherical mapping further.In addition,this paper uses SIFT algorithm to match the feature points of the corrected image and uses RANSAC algorithm to realize the purification and montage of the feature points.The model could accomplish the distortion correction and seamless fusion of the images better.Finally the establishment, the coding and the drive installation of the Android system are done on the development board.
Keywords/Search Tags:Fisheye Image, Distortion Correction, Image Stitching, Android, ARM
PDF Full Text Request
Related items