Title :
Motion video adaptive quantization in the transform domain
Author :
Gonzales, C.A. ; Viscito, E.
Author_Institution :
IBM Thomas J. Watson Res. Center, Yorktown Heights, NY, USA
fDate :
12/1/1991 12:00:00 AM
Abstract :
The authors developed a minimax adaptive quantization algorithm that operates in the discrete cosine transform domain conforming to the Moving Picture Experts Group (MPEG) standard. The algorithm is designed to optimize image quality by adapting a quantizer scaling factor to the local characteristics of the video pictures while preserving a constraint on the average output bit rate. The algorithm is well suited for real-time encoder implementations of current video compression standards, such as MPEG and H.261
Keywords :
data compression; encoding; picture processing; television standards; transforms; video signals; H.261; MPEG standard; Moving Picture Experts Group; average output bit rate; discrete cosine transform; image quality; local characteristics; minimax adaptive quantization algorithm; motion video adaptive quantisation; quantizer scaling factor; real-time encoder; transform domain; video compression standards; video pictures; Algorithm design and analysis; Constraint optimization; Discrete cosine transforms; Discrete transforms; MPEG standards; Minimax techniques; Quantization; Standards development; Transform coding; Video compression;
Journal_Title :
Circuits and Systems for Video Technology, IEEE Transactions on
