The Complexity Of The Context Of Multi-plate Image Segmentation

The vehicle license plate recognition(LPR) system has a wide application in the electronic charging system, assess control system and traffic watching area and so on. As an important part of the vehicle license plate recognition, the vehicle license plate division(LPD) plays a vital role in systematic precision. Present method about LPD is mainly for vehicle image with single car. Sometimes multiple cars will appear in the vehicle image, what's more, the vehicle image perhaps is with complex scenes too. To this kind of situation, a technology for vehicle license plates division from an image with multiple cars and complex scenes is provided. The technology can divide the area in which the car plate is included as small as possible from an image with multiple cars and complex scenes well and truly. So it means much to the development of multiple vehicle license plate recognition.In this thesis, current research status and application of LPR are introduced, the developing and defect about LPD are discussed. And the key techniques including wiping off the complex scenes, multiple vehicle license plates location and division are expatiated.For the complex scenes'wiping off, firstly, the algorithm of feature fuzzy matching is applied to wipe off most of the still and luminous insensitive scenes, then, a kind of area connecting technology is used to wipe off the rest scenes. At last, the vehicle body area which has been mismatched will be repaired. So the vehicle body can be divided from the scenes quickly and exactly.Multiple vehicle license plates location and division is vital in the technology for vehicle license plates division from an image with multiple cars and complex scenes. Because of the abundant upright texture characteristics of the plate area, the algorithm of improved image difference and the algorithm of finding the lowest error limit data are employed as first disposal. Then, a method based on finding the characteristics points is designed to divide the potential license areas. Finally, the morphological approach and projection method for shearing edge are used to finish each suspicious single plate division. At last, each of those suspicious single license areas is checked...