A fast mode decision algorithm applied to Coarse-Grain quality Scalable Video Coding

• Low possible modes are excluded based on spatial and inter-layer correlations.
• We study on residual coefficients and possibilities of mode to early terminate.
• We study on inter-layer correlation and possibilities of mode to early terminate.
• We study on spatial correlation and possibilities of mode to early terminate.

A fast mode decision algorithm is proposed for a Coarse-Grain Scalable (CGS) video encoder based on the encoding characteristics of quality Scalable Video Coding (SVC). First, candidate modes and coding orders are predicted, based on inter-layer and spatial correlations. Three early termination methods are then proposed based on CGS encoding structure. Finally, all candidate modes are checked sequentially, according to their predicted order with three early termination conditions, to improve the coding speed. Experimental results have demonstrated that the proposed algorithm could reduce the encoding time by an average of 84.39%, with negligible coding efficiency losses.