Priority-Processing for Optimized Real-Time Performance with Limited Processing Resources

For future multi-media-systems (MMS) a clear trend can be observed which leads from dedicated hardware to signal processing in software. A problem is to guarantee real-time performance on programmable platforms with varying computational resources. Scalable video algorithms (SVA) which adapt automatically their output quality to the available resources represent thereby a novel concept. We propose the principle of "Priority-Processing" as an efficient way for self-adapting video algorithms to the available platform resources. An SVA for sharpness enhancement has been implemented and tested. Even at varying platform resources real-time processing with very good output quality has been achieved.