Satellite Cloud Image Enhancement by Genetic Algorithm with Fuzzy Technique

In order to efficiently enhance contrast of the satellite cloud image, an efficient satellite cloud image enhancing method is proposed. Most of existing satellite cloud image enhancing methods do not consider uncertainties of the image, that is to say, fuzziness in the satellite cloud image processing. A new kind of image measure function is proposed by combing fuzzy theory with other some measures. We use it as the overall fitness function of genetic algorithm to adaptively optimize the nonlinear transform parameters in the nonlinear transform function. Thus an optimal gray transformation curve is obtained to enhance the region of interest for a satellite cloud image. Typhoon cloud images are used to verb the proposed method. Experimental results show that the proposed method has higher adaptability and intelligence. The proposed method is better than the classical image enhancement methods in overall performance. Enhanced typhoon cloud images by the proposed method can be used to accurately locate the center and predict intensity of the typhoon.