A Complete Video Coding Chain Based on Multi-Dimensional Discrete Cosine Transform
MetadataShow full item record
The paper deals with a video compression method based on the multi-dimensional discrete cosine transform. In the text, the encoder and decoder architectures including the deﬁnitions of all mathematical operations like the forward and inverse 3-D DCT, quantization and thresholding are presented. According to the particular number of currently processed pictures, the new quantization tables and entropy code dictionaries are proposed in the paper. The practical properties of the 3-D DCT coding chain compared with the modern video compression methods (such as H.264 and WebM) and the computing complexity are presented as well. It will be proved the best compress properties could be achieved by complex H.264 codec. On the other hand the computing complexity - especially on the encoding side - is lower for the 3-D DCT method.
KeywordsVideo signal processing, data compression, discrete cosine transform, multidimensional coding, digital signal processors.
Document typePeer reviewed
Document versionFinal PDF
SourceRadioengineering. 2010, vol. 19, č. 3, s. 421-428. ISSN 1210-2512
- 2010/3