Spatial scalable HDTV coding

In this paper, we discuss the coding strategy when a two-layer digital HDTV service is considered. We propose to use-three different strategies in performing compression to yield the best quality based on the respective requirements on bandwidth, picture quality and efficiency. These three modes are as follows. Strategy A: if the bandwidth for the enhancement layer is not sufficient to achieve high quality, the base layer encoder should perform the compression at a reduced spatial resolution using the combined bandwidth. The top layer video signal will be obtained from upsampling the base layer. Strategy B: if the bandwidth is sufficient to support quality requirements at both layers, the typical two-layer spatial scalable coding scheme is used. Strategy C: if the quality requirement at the top layer is critical, a single layer compression at a full spatial resolution is used. To receive the low resolution signal, certain bit stream scaling can be used to get a usable quality signal. A theoretical qualitative analysis gives an insight of our experimental results. Such a strategy is useful in designing future HDTV services for television sets with different capabilities, resolutions and complexities. The results are also applicable to the problem of HDTV/standard TV compatibility