Abstract
AbstractRandom access video compression is mostly implemented without any reduction of temporal redundancy. Standard video compression systems like MPEG (1,2 and 4) are heavily based on motion compensation, which to some extent makes random access at single frame level impossible. We present a method for near random access video compression of low-motion video that is based on the discrete cosine transform and vector quantization and refine this system using weighted finite automata while keeping the random access property and using some reduction of temporal redundancy.KeywordsCompression RatioDiscrete Cosine TransformRandom AccessImage CompressionVector QuantizationThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
