Generation of optimal random fields for image segmentation using fuzzy multi-region technique

This paper presents a competent and optimal technique that can be used to divide the object and background pixels of the image to get better quality of low contrast images by using generation of Optimal Random Fields for Image Segmentation using Fuzzy Multi-region Technique. This innovative method contains two steps: 1) FGFCM is used to get the optimal mean value for the clustering with usual selection of controlled parameters to keep away from the fuzziness in the image. FGFCM exploits a local similarity measure that combines both spatial and gray level image information. 2) This FGFCM output is given to Manifold learning method to advance the contrast of the image. This method gives enhanced results than Manifold learning method. The graphs and tables of values proved that the proposed method is superior.