Image compression is a type of data compression applied to images. The objective of image compression is to reduce the cost for storage or transmission. Image compression is associated with removing redundant information of image data. Image storage is required for several purposes like document, medical images, etc. In this paper, an edge preserving image compression algorithm based on an unsupervised competitive neural network called weighted centroid neural network (WCNN), is implemented and compared to the other algorithms. The WCNN algorithm allots more representative vectors from the edges of the image than the interior of the image thus helping in better edge preservation of the reconstructed image. After experimenting with the cluster count it is evident that with the increase in the number of cluster the quality of the picture is improved, which is the expected behavior as more clusters leads to more representational vectors.