Manjinder Kaur, Navjot Kaur, Harkamaldeep Singh
In the presented work, a modified k-means clustering is proposed. It adapts itself according to the image based on color based clustering. The no. of clusters using the color features are computed based on histogram analysis in gray format. The peak of the histogram is the main source of computation of no. of colors in the image and based on the same, the image data are clustered.