Spatial scalability of multiple ROIs in scalable video coding

In this paper, we propose a new functionality to Scalable Video Coding, which is the support of multiple ROIs. SVC is targeted at the flexible extraction of some bitstream from the original SVC bitstream, and it is discussed in the MPEG committee for standardization. Region of interest (ROI) is an area that is semantically important to a particular user. It is expected that a bitstream that contains the ROI can be extracted without any transcoding operations. In many cases, the user may want to see more than one ROIs at the same time. The existence of multiple ROIs results in some difficulties in extracting the bitstream containing more than one ROI. In this paper, we present solutions to address these difficulties.