Lossless coding using variable block-size adaptive prediction optimized for each image

This paper proposes an efficient lossless coding scheme for still images. The scheme employs a block-adaptive prediction technique to remove spatial redundancy in a given image. The resulting prediction errors are encoded using context-adaptive arithmetic coding. Several coding parameters, which must be sent to a decoder as side information, are iteratively optimized for each image so that the number of coding bits including the side information can have a minimum. Moreover, quadtree-based variable block-size partitioning is introduced into the above adaptive prediction technique. Experimental results show that the proposed coding scheme attains the best coding performance among the current state-of-the-art lossless coding schemes.