Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Selective low bit-rate image compression using wavelet transform

: Santos Conde, J.E.; Niederholz, J.; Hosticka, B.J.

Hamza, M.H.:
Signal processing and communications. Proceedings of the IASTED International Conference
Anaheim: IASTED/ACTA Press, 2000
ISBN: 0-88986-302-4
pp.195-203 : Lit.
International Conference on Signal Processing and Communications (SPC) <2000, Marbella>
Conference Paper
Fraunhofer IMS ()
image processing; Peano-Hilbert-Scan; wavelet; compression; JPEG; Huffman; Bildverarbeitung; Kompression; wavelet-Transformation

A novel image compression approach reaching very low bit-rates by utilization of region of interest information in the transform domain of the discrete wavelet transform (DWT) is introduced. The content-dependent image coding scheme yields an improved quality in the regions of interest while the background regions distant from the center of interest are smooth. Coding distortions are visible for very low bit-rates outside the regions of interest which is tolerable for the purpose of content dependent image compression. The significance of certain image regions can be selected interactively or determined by appropriate image segmentation schemes. Qualitative results are presented for several compression ratios and a comparison with JPEG compression is made.