MC/DCT/HVS/CVQ를 이용한 동영상 압축

Motion Video Compression using MC/DCT/HVS/CVQ

A motion video compression algorithm based on the MC(Motion Compensation) and CVQ(Classified Vector Quantization) in the DCT(Discrete Cosine Transform) domain which takes account of the properties of HVS(Human Visual System) is proposed in this paper. Temporal redundancy is eliminated using ME (Motion Estimation) and MC. Spatial redundancy is eliminated using DCT. Prediction error blocks are transformed using the DCT. These transformed blocks are classified into eight classes based on their structural properties and then subdivided into subvectors depending on the DCT coefficients statistics and the properies of HVS. The subvectors are vector quantized.