Colour image segmentation using fuzzy clustering techniques and competitive neural network

This paper explains the task of segmenting any given colour image using fuzzy clustering algorithms and competitive neural network. The fuzzy clustering algorithms used are Fuzzy C means algorithm, Possibilistic Fuzzy C means. Image segmentation is the process of dividing the pixels into homogeneous classes or clusters so that items in the same class are as similar as possible and items in different classes are as dissimilar as possible. The most basic attribute for image segmentation is the luminance amplitude for a monochrome image and colour components for a colour image. Since there are more than 16 million colours available in any image and it is difficult to analyse the image on all of its colours, the likely colours are grouped together by means of image segmentation. For that purpose soft computing techniques namely Fuzzy C means algorithm (FCM), Possibilistic Fuzzy C means algorithm (PFCM) and competitive neural network (CNN) have been used. A self-estimation algorithm has been developed for determining the number of clusters. The images segmented by these three soft computing techniques are compared using image quality metrics: peak signal to noise ratio (PSNR) and compression ratio. The time taken for image segmentation is also used as a comparison parameter. The techniques have been tested with images of different size and resolution and the results obtained by CNN are proven to be better than the fuzzy clustering technique.