GAUSSIAN MIXTURE MODEL BASED CONTRAST ENHANCEMENT

Abstract

Image enhancement is the process of applying transformation on a digital image to get a better picture. As such there is no way to measure the enhancement of an image except the human perception, however a certain characteristics are used t o determine whether the image is suitable for a particular application. Gaussian mixture model based contrast enhancement serves as a powerful algorithm for image enhancement. The main idea behind the algorithm is to estimate a Gaussian mixture model from the image histogram. . It consists of three phases- First to estimate the Gaussian parameters by using previously mentioned methods and obtain mean, Gaussian boundary , variance and scaling factor. In the second phase, the gamma is calculated using the these parameters for each Gaussian territory. Finally, the calculated gamma is applied through an intensity transformation.