Research And Implementation On Recognition Algorithm Of Blurred Images Of Vehicles And License Plate

With economic development, sharply increased number of vehicles, traffic problems havebecome more prominent, and how to obtain an accurate traffic intelligent transportation systemshas become a hot area of research, by increasing attention. 320 works to the city publicsecurity traffic control of vehicles, vehicle license, especially fuzzy vehicle license platerecognition system to be intelligent monitoring system, traffic and one of the core researchprojects.Blurred image vehicle detection and vehicle license plate recognition system is theintelligent transportation and key technologies of intelligent monitoring. In this paper, the actualcity road monitoring the current poor image quality, focuses on the fuzzy vehicle detection andlicense plate recognition, and license plate recognition algorithm based on the video sequences isproposed using image enhancement algorithm can solve the blurred image vehicle detection andvehicle license plate recognition.Firstly, on fuzzy image pre-processing stage, Richardson-Lucy algorithm, the Wienerfiltering methods of the advantages and disadvantages of analysis, the final, sequence of videoimage enhancement using modified Z transform algorithm and discrete recursive algorithm, theimage restoration, to the follow-up in the process of license plate recognition to meet the basicrequirements. Then the vehicle detection, moving target detection and tracking of some methodsinclude methods based on optical flow field, based on frame difference method and the methodbased on background difference. In the character segmentation, the analysis of the license imageand normalization with the standard geometric correction of the various algorithms, the finalsegmentation using connected region, to achieve the character segmentation. In characterrecognition, the analysis and comparison of the commonly used method for character recognitionis proposed based on structural features and statistical features of the template matching method.Experimental results show that this approach is effective.