Title :
Visibility thresholds for quantization distortion in JPEG2000
Author :
Oh, Han ; Bilgin, Ali ; Marcellin, Michael W.
Author_Institution :
Dept. of Electr. & Comput. Eng., Univ. of Arizona, Tucson, AZ, USA
Abstract :
Visibility thresholds play an important role in finding appropriate quantization step sizes in image and video compression systems. In this paper we present a new method of measuring visibility thresholds for quantization distortion in JPEG2000. A quantization distortion model for each subband is developed based on the statistical characteristics of wavelet coefficients and the dead-zone quantizer of JPEG2000. Thresholds obtained from these quantization distortion models are image-adaptive and ensure visually lossless coding with low computational burden.
Keywords :
data compression; distortion; quantisation (signal); statistical analysis; video coding; visibility; wavelet transforms; JPEG2000; dead-zone quantizer; image compression system; image-adaptive model; quantization distortion model; statistical characteristics; video compression system; visibility threshold; visually lossless coding; wavelet coefficients; Discrete cosine transforms; Discrete wavelet transforms; Distortion measurement; Frequency; Humans; Image coding; Quantization; Transform coding; Video compression; Wavelet coefficients; contrast sensitivity function; human visual system; visibility threshold; visually lossless coding;
Conference_Titel :
Quality of Multimedia Experience, 2009. QoMEx 2009. International Workshop on
Conference_Location :
San Diego, CA
Print_ISBN :
978-1-4244-4370-3
Electronic_ISBN :
978-1-4244-4370-3
DOI :
10.1109/QOMEX.2009.5246948
