Binary-tree recursive motion estimation for video coding

Binary tree predictive coding (BTPC) is an efficient general-purpose still-image compression scheme, competitive with JPEG for natural image coding and with GIF for graphics. We report the extension of BTPC to video compression using motion estimation and compensation techniques which are simple, efficient, non-linear and predictive. The new method, binary tree recursive motion estimation and coding (BTRMEC), exploits the hierarchical structure of BTPC to give progressively refined motion estimates for increasing numbers of pels. Compression results for BTRMEC are compared against conventional block-based motion compensated coding as used in MPEG (i.e. with variable-length coding of motion vectors, and DCT/runlength coding of residues). They show that both BTRMEC and BTPC adapted to video coding are efficient methods to code video sequences