Image compression based on exploiting similarities in a group of pictures

摘要

A lot of images and photos are being produced today using our personal devices such as cameras, smart phones and tablets. Taking many shots is becoming more frequent due to costless processing of pictures production. In this paper, we are investigating compression of a group of similar photos or images all together instead of separate compressions. In our technique, we considered applying blurring over the original images in order to gain better compression ratio. Moreover, subimages are classified to form a non redundant codebook that helps producing index file with smaller size than the original images. Experimental results showed competitive compression achievements compared to peer techniques such as JPEG and PNG.