Lossless Compression of Images

Abstract

Lossless Image Compression requires achieving in both efficient saving of an image and reconstruction of the original image without data loss while decoding. In this project, Huffman Coding will be applied to reach this target. There will be two approaches to reduce the image size. First of all, prediction will be used to calculate a predicted image. It will be subtracted from the original image to get the residual image. The predicted image is expected to be close but not equal to the original image so as their difference, the residual image, is small. Then the residual image will be sent to the Huffman encoder which uses variable-length code representation in order to further reduce the information needed to store each pixel. After encoding, the original image can be reconstructed from the residual image without loss in the decoder with similar approach to that of the reverse of the encoder. The compression ratio achieved in this project is about 1.43 of the original image.