Title :
A low-complexity postprocessing algorithm for block based video coding
Author :
Maddikonda, Ravikanth ; Adhikary, Tushar K.
Author_Institution :
Emuzed India Pvt. Ltd., India
Abstract :
This paper presents a low-complexity algorithm to remove quantization effects in low bit rate video coding, such as blocking artifacts and ringing noise. The postprocessing algorithms presented in literature and MPEG-4 standard are of high computational complexity and apply low pass filtering twice to remove the two quantization effects mentioned above. The proposed algorithm uses the information contained in the compressed bit stream for processing. In this method, the deblocking and deringing operations are combined effectively and filtering is done only once to remove both the noises. This results in reduction of computational complexity. The efficiency of the proposed algorithm is compared with MPEG-4 postprocessing algorithm in terms of computational complexity and peak signal to noise ratio for several video sequences.
Keywords :
computational complexity; data compression; image denoising; image sequences; quantisation (signal); video coding; block based video coding; blocking artifacts; compressed bit stream; computational complexity reduction; low bit rate video coding; low complexity postprocessing algorithm; quantization effect removal; ringing noise; video sequence; Approximation algorithms; Computational complexity; Filtering; PSNR; Quantization (signal); Transform coding; Video coding;
Conference_Titel :
Signal Processing Conference, 2002 11th European
Conference_Location :
Toulouse