Image Compression via Dithering for Enhanced Images

The problems of image storage and transfer are prevalent until today. For several decades now, image compression has been explored with the intent of achieving high compression rates in conjunction with superior image quality. In this paper, a lossy image compression algorithm is proposed that is based on dithering an image using a modified version of the standard error diffusion algorithm. Prior to dithering, the image is first enhanced in the frequency domain. Results indicate that not only is the goal of image compression realized but also the decompressed image promotes retention of fine detail with less random distribution of pixels around object edges. Experiments were conducted on several types of images using both a subjective as well as an objective approach. The compression achieved in terms of the reduction of gray levels is significant and the superior quality of the decompressed image using the method proposed is evident.