License Plate Recognition Under The Haze Weather

Image enhancement has a lot of achievements in various fields with the development and progress of science and technology in recent years. For the reason that the fog and haze weathers to appear frequently in northern China, the processing of the haze image which obtained from the imaging equipment become a problem to be solved urgently.Ⅰ have done the research on enhancement processing of haze image which presented by domestic and foreign scholars on the current situation. I have also done the research on haze image enhancement effect of different algorithms. This paper focuses on the research of retinex theory and the theory of color constancy. Retinex algorithm is the core of a novel image enhancement theory; it can be used for the treatment of fog and haze image. This paper studies the single scale retinex algorithm and multicultural scale retinex algorithm, the two kinds of algorithm in the fog and haze image enhancement application. The single scale SSR algorithm which improved based on the traditional logarithmic function presents a new image enhancement gray adjustment fitting function. The fitting function which meets the human visual can be limited to the brightness change, at the same time; it can guarantee that the image brightness is not beyond its scope of table. These algorithms are compared with each other by different angles, such as the visual angle of the processing results and algorithm of objective parameters objective angle on the treatment effect on the fog and haze image. Then, the paper compares the running time of their computational complexity and algorithm.The fog and haze licenses plate image is improved on image processing, through the preprocessing, license plate location, license plate segmentation, character recognition module operation, access to the final treatment outcome.