A principle of adaptively grouping frames on lossless medical video compression using ideal cross-point regions

Abstract

Grouping frames of a video into groups of pictures (GOP) is an elementary step in video compression to get high compression ratio. In medical field, videos as outputs of some appliances are used for many purposes: for future treatment or for remote diagnosis …, so they must be compressed by using lossless compression methods. This paper proposes a new method to group frames of a video into GOPs based on the theory of ideal cross-point regions. We find the relation between the objects in consecutive frames with the change of the cross-point regions. Experimental results show that the proposed method are effective both for videos which have few or many objects, especially for abrupt change videos.