Contrast enhancement based on fuzzy clustering

The gray level transformation combining with local standard deviation is an effective and efficient way to enhance the contrast of gray images without much calculating burden. Conventional algorithms always calculate pixel´s local standard deviation in a rectangular area centered to itself, which fail to take the natural traits of the image´s content into consideration. The algorithm we propose uses fuzzy clustering to cluster pixels into different types, hence to extract the feature of the image´s content. And modify the pixel combing the standard deviation of the cluster it belongs to with the global one, which enables us to more effectively enhance the image´s contrast. Experimental results show that the proposed method can improve image´s contrast more significantly than the conventional gray level transformation algorithm.