Research On The Application Of Digital Image Processing In Vehicle License Plate Recognition

With the continuous development of economy and improvement of people's living standard, an increasing demand for transportation requires application of intelligent transportation systems. Automatic Vehicle Recognition System as a branch of an intelligent transportation system plays a valuable role in practices of car park management system, public security, traffic management and in related military sector. In a technical perspective, the license plate recognition system is a complex system which involves in many areas varying from digital image processing to pattern recognition.This paper expounds on four techniques in license plate recognition system, which are image pre-processing in license plate recognition system, license plate locating technology, character segmentation and character recognition, and methods used in studying each including (1) vehicle license image gray-scale transformation method image, binary methods and smoothing methods on study of pre-processing techniques. (2) morphology-based license plate location method on study of license plate positioning technology. (3) a method combined with global threshold segmentation and vertical projection on study of license plate character segmentation algorithm. (4) a method combined with cluster analysis and BP neural network on the study of license plate character recognitionA license plate recognition system is developed based on this theoretical study and with Visual C++6.0 as a tool. Experimental data shows that the system functions in a more accurate way of locating, segmentation and identification license plate and license plate characters from which we draw a conclusion that a segmentation positioning technology combined with a number of ways can improve overall system performance.