LAUSR

Abstract This paper investigates a coding method for medical video compression. The described technique has advantages of providing a higher visual quality of video coding, and efficiently reducing spatial redundancy. Firstly the bandelet transform, also called a non-separable transform, is studied in order to allow an efficient detection of the different complex geometries found in video. After describing the lifting operation in the bandelet transform, we reduced the artifactual effect caused by the quad-tree decomposition step in order to enhance the visual quality of the medical video sequence. Finally, the efficiency of the proposed method on medical video is tested and evaluated by means of a set of objective measurement parameters.