Improving screen content coding in HEVC by transform skipping

Screen content is nowadays a part of numerous applications - from desktop sharing to broadcasting. It consists of both camera captured content and computer generated content such as text and graphics. These two types of content have different properties requiring different processing and compression techniques. However, it is often required to compress such content with standard video coding solutions. In this paper a low-cost solution for improved screen content coding based on the upcoming video coding standard HEVC is presented. It includes specific intra and inter coding solutions that skip some of the common video coding methods, in this case transforms, enhancing the compression of screen content. Additional signalling and signal-level adjustment methods are introduced. Savings of up to 30% of the bit-rate are observed for intra coding and up to 25% for inter coding of screen content. Modest gains of up to 3% are observed for content that consists of both camera captured content and graphics.