Publication Details

Vectorization and Parallelization of 2-D Wavelet Lifting

BAŘINA David and ZEMČÍK Pavel. Vectorization and Parallelization of 2-D Wavelet Lifting. Journal of Real-Time Image Processing, vol. 15, no. 2, pp. 349-361. ISSN 1861-8200.
Czech title
Vektorizace a paralelizace 2D diskrétní vlnkové transformace pomocí schématu lifting
Type
journal article
Language
english
Authors
Keywords
Discrete wavelet transforms, Image processing
Abstract
The article presents several novel SIMD-vectorized algorithms of 2-D discrete wavelet transform. For all of the platforms used in the tests, these algorithms are significantly faster than other known methods.

The final publication is available at Springer via http://dx.doi.org/10.1007/s11554-015-0486-6.

Annotation
With the start of the widespread use of discrete wavelet transform in image processing, the need for its efficient implementation is becoming increasingly more important. This work presents several novel SIMD-vectorized algorithms of 2-D discrete wavelet transform, using a lifting scheme. At the beginning, a stand-alone core of an already known single-loop approach is extracted. This core is further simplified by an appropriate reorganization of operations. Furthermore, the influence of the CPU cache on a 2-D processing order is examined. Finally, SIMD-vectorizations and parallelizations of the proposed approaches are evaluated. The best of the proposed algorithms scale almost linearly with the number of threads. For all of the platforms used in the tests, these algorithms are significantly faster than other known methods, as shown in the experimental sections of the paper.
Published
2015
Pages
349-361
Journal
Journal of Real-Time Image Processing, vol. 15, no. 2, ISSN 1861-8200
Publisher
Springer Verlag
DOI
BibTeX
@ARTICLE{FITPUB10801,
   author = "David Ba\v{r}ina and Pavel Zem\v{c}\'{i}k",
   title = "Vectorization and Parallelization of 2-D Wavelet Lifting",
   pages = "349--361",
   journal = "Journal of Real-Time Image Processing",
   volume = 15,
   number = 2,
   year = 2015,
   ISSN = "1861-8200",
   doi = "10.1007/s11554-015-0486-6",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/10801"
}
Files
Back to top