Evaluation of Classification in Markov Random Field (MRF) Segmented Images in the Presence of Gaussian Noise

Abstract

The major objective of computer vision is to enable the machine to understand the world of visual observation through the processing of digital signals. Such an interpretation for the machine is done by extracting useful information from the digital signals and performing complex computation. Energy minimization in image segmentation is a distinguished approach in computer vision. One of the best suited model for image segmentation is Markov Random Fields (MRF). In this paper segmentation is carried out with the help of MRF which estimates the classification level in each region of the image in the presence of Gaussian Noise