Wavelet image coding by dilation-run algorithm

• Published in: 2004 International Conference on Image Processing Vol. 5; pp. 3165 - 3168 Vol. 5
• Main Authors: WU, Zheng, HE, Mingyi
• Format: Conference Proceeding
• Language: English
• Published: Piscataway NJ IEEE 2004
• Subjects: Applied sciences; Artificial intelligence; Clustering algorithms; Computer science; control theory; systems; Data mining; Discrete wavelet transforms; Exact sciences and technology; Helium; Image analysis; Image coding; Morphology; Pattern recognition. Digital image processing. Computational geometry; Wavelet analysis; Wavelet coefficients; Wavelet transforms; Cluster analysis; Image analysis; Image coding; Image processing; Wavelet transformation; Morphology; Classification; Positional information
• ISBN: 0780385543; 9780780385542
• ISSN: 1522-4880
• DOI: 10.1109/ICIP.2004.1421785

This paper presents a novel wavelet image coder, dilation-run algorithm, which provides an embedded coder based on bit-plane coding technique. At each bit-plane, morphological dilation is used to extract and encode the clustering significant wavelet coefficients, and a run-length coding method is used to encode the positional information of each cluster. The experiment results show that our coder outperforms the zerotree coder SPIHT and is competitive with the morphology coders MRWD and SLCCA. For wavelet images with strong clustering feature, the new coder outperforms both the morphology coders above.