Efficient Enhancement for Spatial Scalable Video Coding Transmission

Scalable Video Coding (SVC) is an international standard technique for video compression. It is an extension of H.264 Advanced Video Coding (AVC). In the encoding of video streams by SVC, it is suitable to employ the macroblock (MB) mode because it affords superior coding efficiency. However, the exhaustive mode decision technique that is usually used for SVC increases the computational complexity, resulting in a longer encoding time (ET). Many other algorithms were proposed to solve this problem with imperfection of increasing transmission time (TT) across the network. To minimize the ET and TT, this paper introduces four efficient algorithms based on spatial scalability. The algorithms utilize the mode-distribution correlation between the base layer (BL) and enhancement layers (ELs) and interpolation between the EL frames. The proposed algorithms are of two categories. Those of the first category are based on interlayer residual SVC spatial scalability. They employ two methods, namely, interlayer interpolation (ILIP) and the interlayer base mode (ILBM) method, and enable ET and TT savings of up to 69.3% and 83.6%, respectively. The algorithms of the second category are based on full-search SVC spatial scalability. They utilize two methods, namely, full interpolation (FIP) and the full-base mode (FBM) method, and enable ET and TT savings of up to 55.3% and 76.6%, respectively.