14 November 2007 Image compression with directional lifting on separated sections
Author Affiliations +
Proceedings Volume 6790, MIPPR 2007: Remote Sensing and GIS Data Processing and Applications; and Innovative Multispectral Technology and Applications; 67905I (2007) https://doi.org/10.1117/12.752509
Event: International Symposium on Multispectral Image Processing and Pattern Recognition, 2007, Wuhan, China
Abstract
A novel image compression scheme is presented that the directional sections are separated and transformed differently from the rest of image. The discrete directions of anisotropic pixels are calculated and then grouped to compact directional sections. One dimensional (1-D) adaptive directional lifting is continuously applied along orientations of direction sections other than applying 1-D wavelet transform alternately in two dimensions for the whole image. For the rest sections, 2-D adaptive lifting filters are applied according to pixels' positions. Our single embedded coding stream can be truncated exactly for any bit rate. Experiments have showed that large coefficients can be significantly reduced along directional sections by our transform which makes energy more compact than traditional wavelet transform. Though rate-distortion (R-D) optimization isn't exploited, the PSNR is still comparable to that of JPEG-2000 with 9/7 filters at high bit rates. And at low bit rates, the visual quality is better than that of JPEG-2000 for along directional sections both blurring and ringing artifacts can be avoided and edge preservation is good.
© (2007) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Jieying Zhu, Jieying Zhu, Nengchao Wang, Nengchao Wang, } "Image compression with directional lifting on separated sections", Proc. SPIE 6790, MIPPR 2007: Remote Sensing and GIS Data Processing and Applications; and Innovative Multispectral Technology and Applications, 67905I (14 November 2007); doi: 10.1117/12.752509; https://doi.org/10.1117/12.752509
PROCEEDINGS
9 PAGES


SHARE
RELATED CONTENT

Wavelet-based image compression using fixed residual value
Proceedings of SPIE (December 27 2000)
Embedded wavelet coding of arbitrarily shaped objects
Proceedings of SPIE (May 29 2000)
Centralized and interactive compression of multiview images
Proceedings of SPIE (September 23 2011)
Adaptive boxcar/wavelet transform
Proceedings of SPIE (January 18 2009)

Back to Top