Font Size: a A A

License Plate Recognition System Of License Plate Location And Tilt Correction And Character Segmentation Technology Research

Posted on:2014-02-02Degree:MasterType:Thesis
Country:ChinaCandidate:X G DuFull Text:PDF
GTID:2248330395992131Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
The technology of License Plate Recognition is based on the intelligentidentification system of the technology of digital image processing, patternrecognition, computer vision and so on. Eventually,this technology which makes thecomputerized monitoring and management vehicle become a reality is An importantresearch topic.In brief,the system of PLR must be to meet the following requirements:thefirst,it is the robustness of the system of PLR.In any case,the system can reliablyand normally run,and keep a high recognition rate;the second,it is the real-time ofthe system of PLR. No matter in the car still or high speed operation conditions,thesystem can timely give the right result of recognition,meet the real time requirements.To these questions of shortcomings and insufficiency of the past algorithm ofvehicle plate recognition system,therefore,this paper will deeply analysis and researchthese key links of License Plate Location, Image Preprocessing, Slant Correction,Character Segmentation of the LPR,meanwhile,design a complete license platerecognition system carried out in MATLAB simulation.The part of License Plate Location,this paper puts forward a fast and accurateLicense plate location algorithm based on the characteristics of RGB Color Space ofLicense Plate area and texture of Characters.Firstly,it is based on the Blue componentof License Plate area to implement the function of preliminary location of LicensePlate.Because the body of License Plate itself may be blue,it necessarily needs tocombine with the texture feature of character area of License Plate to implementaccurate location of License Plate.The part of Image Preprocessing,firstly,it makes a special gray level change onthe location image of License Plate;secondly,it makes a image subtraction operation tohighlight Characters and restrain noise of the background;finally,it makes a the imagebinary operation by the way of OSTU to get a clear and low-noise binary image ofLicense Plate.The part of Slant Correction,,this paper puts forward a fast and effective correction algorithm of mixed slant vehicle License plate.The part of Character Segmentation,to these questions of shortcomings andinsufficiency of weak anti-interference ability and adaptability of the pastalgorithm,this paper puts forward a fast and accurate Character Segmentationalgorithm based on the connected character domain.
Keywords/Search Tags:License Plate Location, Slant Correction, Character Segmentation, Cluster Analysis, connected components labeling of Character
PDF Full Text Request
Related items