1 April 2006 Compression of 3-D echocardiographic images using a modified 3-D set-partitioning-in-hierarchical-trees algorithm based on a 3-D wavelet packet transform
Author Affiliations +
J. of Electronic Imaging, 15(2), 023016 (2006). doi:10.1117/1.2194467
Abstract
An efficient compression strategy is indispensable for 3-D digital echocardiography, which can provide more accurate diagnostic information than 2-D echocardiography without geometric assumption, but unfortunately requires a huge storage space. We describe a new set partitioning in hierarchical trees (SPIHT) algorithm that is based on the wavelet packet transform and use it to compress 3-D echocardiographic images. The new algorithm originates from the well-known SPIHT algorithm that is based on a wavelet transform but outperforms it for the compression of 3-D echocardiographic images. Experimental results are presented to verify the effectiveness of the new SPIHT algorithm. Results show that the new algorithm achieves a very high compression ratio while retaining good quality 3-D echocardiographic images.
Xiyi Hang, Neil L. Greenberg, Yuan F. Zheng, James D. Thomas, "Compression of 3-D echocardiographic images using a modified 3-D set-partitioning-in-hierarchical-trees algorithm based on a 3-D wavelet packet transform," Journal of Electronic Imaging 15(2), 023016 (1 April 2006). http://dx.doi.org/10.1117/1.2194467
JOURNAL ARTICLE
13 PAGES


SHARE
KEYWORDS
Image compression

Wavelets

3D image processing

Lithium

3D video compression

Image segmentation

Wavelet transforms

RELATED CONTENT

Efficient frame interpolation for Wyner-Ziv video coding
Proceedings of SPIE (January 19 2009)
Character recognition using a multistage neural network
Proceedings of SPIE (September 16 1992)
Measurement error analysis of the 3D four-wheel aligner
Proceedings of SPIE (October 10 2013)
Signal-adapted transform coding of sequences
Proceedings of SPIE (October 22 1993)
Multiscale geometric filter based on the wavelet transform
Proceedings of SPIE (October 23 1996)

Back to Top