Integer computation of lossy JPEG2000 compression.
basic_science · Level V
Where this comes from
- Record sourced from PubMed, PMID 21324784.
- Also identified by DOI 10.1109/TIP.2011.2114353.
- No licence information is recorded for this record.
- Because redistribution is not established, this page shows the abstract only. Follow the links below for the full text.
Abstract
In this paper, an integer-based Cohen-Daubechies-Feauvea (CDF) 9/7 wavelet transform as well as an integer quantization method used in a lossy JPEG2000 compression engine is presented. The conjunction of both an integer transform and quantization step allows for a complete integer computation of lossy JPEG2000 compression. The lossy method of compression utilizes the CDF 9/7 wavelet filter, which transforms integer input pixel values into floating-point wavelet coefficients that are then quantized back into integers and finally compressed by the embedded block coding with optimal truncation tier-1 encoder. Integer computation of JPEG2000 allows a reduction in computational complexity of the wavelet transform as well as ease of implementation in embedded systems for higher computational performance. The results of the integer computation show an equivalent rate/distortion curve to the JasPer JPEG2000 compression engine, as well as a 30% reduction in computation time of the wavelet transform and a 56% reduction in computation time of the quantization processing on an average.