Scalable Video Coding Using Hybrid DCT/Wavelets Architectures

This paper proposes the use of wavelet image transformation and polyphase downsampling in scalable video coding. A wavelet-based inter-frame coding solution using the syntax and framework of both MPEG-4 H.264/AVC and it is scalable extension, SVC. In the former codec, redundant slices are employed for coding the high frequency subbands of wavelet transformed imaged. While in the latter codec, the wavelet subbands are arranged into separate Coarse Grain Scalability (CGS) layers. Additionally, the paper proposes the use of a modified polyphase downsampling in applications of scalability and error resiliency. It is shown that the coding efficiency of the proposed solutions is comparable to single layer coding.