Data Compression of Digital Image By Dividing It into Variable Block Sizes

Abstract

In this paper, we introduced a new technique to code the digital images by dividing it into variable block sizes. We choose block in small size for active area (edges) and large one for non-active region (smoothed). After the dividing operation, the discrete cosine transform operates on the different block sizes. Furthermore, many operations will employ to minimize the bit rate.The receiver will reverse all the operations that have been done at the sender and reconstruct the image by using side information that combines with the operations that have been done at the sender. Consequently, this technique shows improvement in the reconstructed image equality by using low bit rate comparing with the available techniques. The reconstructed image is different slightly due to the errors that come from the quantization process, which are impossible to avoid it.