REDUCTION OF REDUNDANT PIXELS OF AN IMAGE USING EFFICIENT KMEAN CLUSTERING ALGORITHM

Clustering means grouping of similar objects which is an important and essential part of Data Mining. Now a day's huge amount of idol data transferring over internet is an image data which consumes large memory size parallel takes more processing time.Similarlyfor transferring high quality image we require high bandwidth. Using some clustering algorithms redundancy pixels problem is reduced. Now in order to retrieve the image with reduction of redundant pixels there is a considerable amount of loss of data of an image. For this problem optimized method is chosen to provide a final image which is comparatively of smaller memory size than the original image and also visually similar to that of an original image. Hence we use K-means clustering algorithm which is essential. But existing K-means algorithm has some problems which include complex computations and also selection of centroid and cluster size should be initially declared. So we use improved kmeans algorithm to generate the clusters with best accuracy and less computational time.