Detecting license plate based on top-hat transform and wavelet transform

In a vehicle license plate identification system, plate region detection is the key step before the final recognition. This paper presents a license plate detection algorithm from complex background based on top-hat transform and wavelet transform. First, the vertical gradient of gray image is calculated and the region of license plate is standed out by using top-hat transformation. Then the horizontal projection of gradient variance is obtained and horizontal candidates are located by using threshold. Multiresolution feature of candidates are extracted by using wavelet transformation in the following. Finally, accurate vertical position of license plate is detected and license plate is extracted from image. In order to show the effectiveness of the proposed method, experiments have been conducted on 360 images taken from various scenes and different conditions. The algorithm can quickly and correctly detect the region of license plate and the license plate detecting rate of success is 98.61%.