DefinePK

Abstract

The increase use of transfer of images from one device to another created a need to compress them without compromising their quality. Machine learning algorithm forms a potential method to compress an image by clustering thus leading to elimination of redundant colours. K-means algorithm is an unsupervised machine learning algorithm which clusters data and hence elected as best method for compression of colours. In this paper we propose compression of an image using k-means algorithm and variation of size of image for different values of k. Moreover comparison of k-means with JPEG and wavelet compression algorithms is performed deriving their advantages, disadvantages and compression ratio.