Font Size: a A A

A Recognition Algorithm Research To Equatorial Guinea Vehicle License Plates

Posted on:2018-04-22Degree:MasterType:Thesis
Country:ChinaCandidate:X P WuFull Text:PDF
GTID:2348330533966420Subject:Engineering
Abstract/Summary:PDF Full Text Request
With the rapid economic and social development,and the improvement of people’s living standards,the number of motor vehicles is also increasing.To improve the management efficiency of the vehicle,to ease traffic pressure on the road,an effective intelligent traffic management system is very necessary.Rapid and accurate identification of the license plate is an essential function of an intelligent traffic management system.An accurate and quick algorithm for identifying the car license plate number from the camera image in public,will greatly improve the convenience of getting the car plate information,thereby enhancing the efficiency of traffic management and traffic peccancy punishment.Vehicle license plate recognition technology is one of the important research topics in computer vision and pattern recognition technology used in intelligent transportation field.This research is the algorithm and technology solutions for License Plate Recognition used in Africa Equatorial Guinea.According to the searching result,I have not found relevant researches devoted to identifying Equatorial Guinea license plates so far.In different countries,the overall process flow of license plate recognition algorithm is common.But the specific design for an algorithm module,need to analyze the applicability of various algorithms based on the actual situation,and then those algorithms would be improved or abandoned.Only in this way,an effective license plate recognition algorithm for Equatorial Guinea can be developed.The processing modules of license plate recognition algorithm in this paper includes: license plate location,character segmentation,character recognition,license plate tracking and result output.For each module,this paper analyzes the existing theory of algorithms and propose a solutions with practical significance:1.The license plate location module: The process that combining a variety of methods such as color distribution,texture analysis,candidate area screening,can position of the plate coarsely,and then the processing steps including morphological processing,tilt correction,relocation and other methods in the rough location area can obtain precise positioning of the plate regional;2.The character segmentation module: Based on general vertical projection segmentation method,a segmentation method was designed specifically for Equatorial Guinea license plate;3.The character recognition module: using BP artificial neural network model,the model has good generalization ability and high recognition accuracy;4.The plate tracking and result output module: Because the processing flow is used to video stream,the license plate detection algorithm and license plate tracking algorithm is integrated in this paper.For the license plate which appears in number of consecutive video frames,the algorithm starts tracking when the license plate occurs,until the plate disappears,then count the recognition results in multiple frames,and obtain the optimal result by the maximum likelihood method.The experiment verify that : the algorithm in this paper does well in detecting,tracking,and recognizing the License Plate appeared in the test video,and the recognition accuracy rate reaches to 93.33%.
Keywords/Search Tags:Equatorial Guinea license plate recognition, color pairs analysis, Hough Transform, BP artificial neural network, CamShift algorithm
PDF Full Text Request
Related items