Research On The Algorithms Of Vehicle License Plate Image Location

As coming along with the development of the highway traffic enterprise in the new century, the requirement of the Intelligent Traffic System (ITS) gets into a new level. Vehicle license plate recognition (VLPR) which synthesizes computer vision and pattern recognition plays an important part in ITS. The license plate recognition can automatic-ally provide vehicles inspecting, validation, registering and alarming, so it can be used in high way management systems, parking charging systems, uptown vehicle management systems and electronic police systems.This paper accomplishes the study of vehicle license characteristic, improves locating algorithm and presents a new method which is the same with DSP. Main contents of the paper consist of four parts: 1. The paper introduces present situation and future development of VLPR, presents the composing of VLPR system and analyzes the character and characteristic of the equipments in the system. 2. Discusses an algorithm for skew remedy based on searching the main direction. The algorithm makes the most of the features of skew image, computes the horizontal and vertical difference of the distorted image and finds the main direction which makes the discriminate function to achieve maximum. 3. Mainly presents license plate location method based on combination of several algorithms. First, this algorithm which is applied to the license plate image after being preprocessed locates the license plate approximately with template matching. An adaptive method of data stream modification which can change the size of template according to current object is adopted. It can avoid the limitations of setting fixed template. An optimization algorithm about template matching is used for increasing the efficiency. Then edge detection based on morphological grads provides better license plate's edge and is propitious to locate. Finally, projection afforded exact location and the license plate location can be got through projection analysis. 4. Simply presents DSP. The system consists of TMS320VC5416, video input processor SAA7113 and a high capacity memory, so it can acquire process and store image. It can satisfy real time and veracity requirements.By the experiment, simulation result indicates that the proposed method has speediness and accuracy to search vehicle license location. It proves that the algorithm has higher feasibility.