A study of Graphics Processing Technologyon Recognition of Vehicles’ Number PlatesA study of Graphics Processing Technology on Recognition of Vehicles’ Number Plates

• Main Authors: chin-te liao, 廖進德
• Other Authors: Ching-Liang Su
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/56401774615690916399

碩士 === 大葉大學 === 工業工程與科技管理學系碩士在職專班 === 95 === The number of transport ownership has surged in recent years, bringing problems as the lack of parking lot, the theft of cars etc. It is patent that the management of cars is an important issue to be solved and the different number plate of each vehicle is a very useful data base. The method shown in this research is to adapt webcam connected to notebooks in order to retrieve the number plates as direct image input in a form of motion graphic. Machine vision technology is the main application. The process is divided into three phrases: video pre-processing, locating of number plates and the separation of characters in motion graphic, and the detection of characters on the number plate. Preprocessing method includes capture of motion graphic, transformation of colors, bi-level quantization, improvement of the quality of number plate images by lowering the unnecessary noise and detection. With the images of characters on number plates strengthened, each character is tracked, extracted and recognized. A database containing all the characteristics and features of every character will be established beforehand in order to facilitate the detection of characters on the number plates by template matching.