Stack-Run Adaptive Wavelet Image Coding
نویسندگان
چکیده
Wavelets are not well-suited to represent oscillatory patterns, a special form of texture. However, results from various image coding methods based on wavelet packets, such as [1], show that they are particularly good in coding images with oscillatory patterns. In this paper, we present an adaptive wavelet transform based image coder which employs a stack-run representation for quantized transform coefficients in order to benefit from the intra-subband redundancies. Our compression algorithm can be divided into four parts: In the first part, an adaptive wavelet packet basis is selected for representing the given image using certain entropy-based cost function. Next, the wavelet packet coefficients are quantized using an optimal scalar quantizer for Laplacian distribution. In the third part, we represent the quantized coefficients with stack-run coding [3] generating a redundant symbol stream. Finally, this symbol stream is entropy coded using a higher order arithmetic coder. The coder variations were based on three different filters: symmlet-8, biorthogonal 17/11, and biorthogonal 9/7 filters. Each of these variations employed three additive cost functions for basis selection: Coifman-Wickerhauser (CW) entropy, energy, and l1-norm. Coding gains achieved by our SRWP coder over both SPIHT [2] and FAWP [1] coders are noticeable, especially for Barbara (1.2dB over SPIHT and 0.4dB over FAWP). Another interesting result is the consistently better performance of the symmlet-8 filters over filters in popular use.
منابع مشابه
Stack-run Adaptive Wavelet Image Compression Using Wavelet Packets
We report on the development of an adaptive wavelet image coder based on stack-run representation of the quantized coefficients. The coder works by selecting an optimal wavelet packet basis for the given image and encoding the quantization indices for significant coefficients and zero runs between coefficients using a 4-ary arithmetic coder. Due to the fact that our coder exploits the redundanc...
متن کاملWavelet Based Image Coding Schemes : A Recent Survey
A variety of new and powerful algorithms have been developed for image compression over the years. Among them the wavelet-based image compression schemes have gained much popularity due to their overlapping nature which reduces the blocking artifacts that are common phenomena in JPEG compression and multiresolution character which leads to superior energy compaction with high quality reconstruc...
متن کاملStack-run image coding
We describe a new image coding approach in which a 4-ary arithmetic coder is used to represent signiicant coeecient values and the lengths of zero runs between coeecients. This algorithm works by raster scanning within subbands, and therefore involves much lower addressing complexity than other algorithms such as zerotree coding that require the creation and maintenance of lists of dependencies...
متن کاملStack-run-end compression for low bit rate color image communication
A new wavelet image coding algorithm was designed for color image compression in this paper. This algorithm utilizes multiary symbol set to represent the meaningful coefficients in the wavelet transform domain which are necessary for the image reconstruction in the respective color channel. The scheme works first by color space conversion, followed by raster scanning the individual subband for ...
متن کامل2 Transform and Encoding Algorithm
The present contribution proposes a new remarkably eecient image compression algorithm for graylevel images based on dyadic wavelet transformation. In order to achieve perfect reconstruction, orthogonal decomposition is applied. Scalar quantization of wavelet coeecients is combined with run-length coding. Code word assignment is performed by semi-adaptive Huuman coding (determined by validity t...
متن کامل