A kind of license plate location based on mathematical morphology and edge detection

In this paper, the license location is adopted by using mathematical morphology and edge detection. Using the erosion operation with the horizontal structure element and the subtraction operation, we cope with the gray image turned into by the inputted color license image in order to obtained the gray image with the prominent license plate characters. Then detect the edge of the gray image by using the sobel operator. Eliminate the noise in the obtained binary image by using the erosion operation with horizontal structure element and then by using the close operation of the mathematical morphology. Some connecting areas are formed. Take two thresholds filter the connecting areas so that the connecting area with the license plate is confirmed. Achieve the accurate location of the license plate by the area preparations and the scanning row and column. The algorithm is carried with 76 color license plate pictures. The results show that the positioning accuracy rate of this algorithm comes up to 96%, and it can meet certain real-time requirements.