Efficient Approach for Image Segmentation Using the Concept of Pixon and Markov Random Field

Segmentation is one of the most complicated procedures in the image processing that has important role in the image analysis. In this paper، using the pixon concept، a segmentation algorithm is proposed based on Markov Random Field (MRF). In the proposed algorithm images are initially modeled using pixonal images، and then the MRF-based segmentation process operates on the pixons to produce the segmented image. The results indicate that the proposed algorithm is superior to the existing image segmentation techniques in terms of segmentation accuracy and speed.