A License Plate Image Enhancement Algorithm under Complicated Illumination

In order to solve the problem caused by the complex illumination of license plate recognition, an image enhancement algorithm under complex illumination based on improved Retinex algorithm is presented. First of all, extract the brightness of the images and then process the brightness component with top-hat transformation, using this method the background of the license plate images can be extracted accurately and then extract the character, attain the purpose that effectively reduces the influence of the complex illumination; next enhance the image details and eliminate the halo at the same time with improved Retinex algorithm and get the final results. Experiments show that for the license plate images in complex light, the method in this paper fully reflects the advantages in both subjective visual effect and objective indicators, proves the feasibility of this method.