Classification based histogram specification framework for image contrast enhancement

In this paper, we propose a novel contrast enhancement technique, referred to as the classification based histogram specification framework (CHSF). CHSF consists of offline training and online enhancement processes. In the offline training process, training images are classified into multiple classes, and then the feature vector that can discriminate the classes is generated and stored along with its most appropriate target histogram. In the online enhancement process, the best matching class is determined by comparing the feature vector of each class and that of the input image, then the corresponding target histogram of the class is adopted for histogram specification (HS). Experimental results show that the proposed method effectively enhances the image contrast by bringing out image details without amplifying noise in flat regions.