Faculty of Information Technology, BUT

Publication Details

CUDA Accelerated Real-time Digital Image Stabilization in a Video Stream

PACURA Dávid and DRAHANSKÝ Martin. CUDA Accelerated Real-time Digital Image Stabilization in a Video Stream. International Journal of Software Engineering and Its Applications, vol. 10, no. 6, pp. 113-124. ISSN 1738-9984.
Czech title
Digitální stabilizace obrazu v reálném času ve videostreamu s využitím akcelerace CUDA
Type
journal article
Language
english
Authors
Pacura Dávid, Ing. (FIT BUT)
Drahanský Martin, prof. Ing., Dipl.-Ing., Ph.D. (DITS FIT BUT)
URL
Keywords
digital image stabilization, CUDA, real-time video stabilization
Abstract
The most important step for successful video processing in computer vision is its stabilization. Often, it is required to process high resolution video in a real-time. In this paper, a new method for a real-time digital image stabilization in a video stream is presented. This method preserves the intended camera motion and exploits computing power of GPGPU by utilizing CUDA programing interface. In order to reduce required computation power, local search windows are used for the correspondence search of consecutive video frames. These windows are further processed using Local Binary Patterns, which enables fast correlation using bitwise XOR. The experiments on video sequences from both car-mounted and hand-held camera have demonstrated the effectiveness of this method. The speed of stabilization designates this method for video preprocessing in real-time applications.
Published
2016
Pages
113-124
Journal
International Journal of Software Engineering and Its Applications, vol. 10, no. 6, ISSN 1738-9984
Publisher
Science & Engineering Research Support Center
DOI
BibTeX
@ARTICLE{FITPUB11183,
   author = "D\'{a}vid Pacura and Martin Drahansk\'{y}",
   title = "CUDA Accelerated Real-time Digital Image Stabilization in a Video Stream",
   pages = "113--124",
   journal = "International Journal of  Software Engineering and Its Applications",
   volume = 10,
   number = 6,
   year = 2016,
   ISSN = "1738-9984",
   doi = "10.14257/ijseia.2016.10.6.10",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/11183"
}
Back to top