Farag I. Y. Elnagahy
Czech Technical University in Prague
Faculty of Electrical Engineering
Department of Telecommunications
16627, Praha 6, Czech Republic
ABSTRACT
This paper proposes an embedded rate scalable wavelet-based image coding algorithm. We introduce a simple and efficient approach for coding the positions and the signs of the wavelet coefficients that will be transmitted as nonzero values in an embedded scheme. This algorithm is based on recursive partitioning of the significant wavelet sub-bands (RPSWS). The proposed algorithm produces a fully embedded bit stream and it encodes the image to exactly the desired bit rate (precise rate control). Some standard test images were used to verify the correctness of the proposed coding algorithm. Their results were compared to the two well-known embedded image coders: embedded zerotree wavelet coder (EZW) and set partitioning in hierarchal trees coder (SPIHT). These two techniques make use of “spatial orientation trees” (SOT). Spatial orientation trees are structures that use quad-tree representations of wavelet coefficients that belong to different sub-bands, but have the same spatial location.
Keywords:
Image compression, Scalable, Embedded
coding, Wavelet sub-band partitioning.