Publikace
Detail publikace
Citace
p. 224-233, Springer, 2012. : GPU Accelerated Real Time Rotation, Scale and Translation Invariant Image Registration Method . International Conference on Image Analysis and Recognition, Lecture Notes in Computer Science, vol. 7324,
PDF ke stažení
Abstrakt
This paper presents highly optimized implementation of image registration method that is invariant to rotation scale and translation. Image registration method using FFT works with comparable accuracy as similar methods proposed in the literature, but practical applications seldom use this technique because of high computational requirement. However, this method is highly parallelizable and offloading it to the commodity graphics cards increases its performance drastically. We are proposing the parallel implementation of FFT based registration method on CUDA and OpenCL. Performance analysis of this implementation suggests that the parallel version can be used for real time image registration even for image size up to 2k x 2k. We have achieved significant speed up of up to 345x on NVIDIA GTX 580 using CUDA and up to 116x on AMD HD 6950 using OpenCL. Comparison of our implementation with other GPU based registration method reveals that our implementation performs better compared to other implementations.
Abstrakt v češtině
This paper presents highly optimized implementation of image registration method that is invariant to rotation scale and translation. Image registration method using FFT works with comparable accuracy as similar methods proposed in the literature, but practical applications seldom use this technique because of high computational requirement. However, this method is highly parallelizable and offloading it to the commodity graphics cards increases its performance drastically. We are proposing the parallel implementation of FFT based registration method on CUDA and OpenCL. Performance analysis of this implementation suggests that the parallel version can be used for real time image registration even for image size up to 2k x 2k. We have achieved significant speed up of up to 345x on NVIDIA GTX 580 using CUDA and up to 116x on AMD HD 6950 using OpenCL. Comparison of our implementation with other GPU based registration method reveals that our implementation performs better compared to other implementations.
Detail publikace
Název: | GPU Accelerated Real Time Rotation, Scale and Translation Invariant Image Registration Method |
---|---|
Autor: | Sudhakar Sah ; Jan Vaněk ; YoungJun Roh ; Ratul Wasnik |
Název - česky: | GPU Accelerated Real Time Rotation, Scale and Translation Invariant Image Registration Method |
Jazyk publikace: | anglicky |
Datum vydání: | 20.6.2012 |
Rok vydání: | 2012 |
Typ publikace: | Stať ve sborníku |
Název knihy: | International Conference on Image Analysis and Recognition |
Svazek: | Lecture Notes in Computer Science |
Číslo vydání: | 7324 |
Strana: | 224 - 233 |
DOI: | 10.1007/978-3-642-31295-3_27 |
ISBN: | 978-3-642-31294-6 |
Nakladatel: | Springer |
Klíčová slova
GPU, Image Registration, CUDA, OpenCL, Object Recognition
Klíčová slova v češtině
GPU, Image Registration, CUDA, OpenCL, Object Recognition
BibTeX
@INPROCEEDINGS{SudhakarSah_2012_GPUAcceleratedReal, author = {Sudhakar Sah and Jan Van\v{e}k and YoungJun Roh and Ratul Wasnik}, title = {GPU Accelerated Real Time Rotation, Scale and Translation Invariant Image Registration Method}, year = {2012}, publisher = {Springer}, volume = {7324}, pages = {224-233}, booktitle = {International Conference on Image Analysis and Recognition}, series = {Lecture Notes in Computer Science}, ISBN = {978-3-642-31294-6}, doi = {10.1007/978-3-642-31295-3_27}, url = {http://www.kky.zcu.cz/en/publications/SudhakarSah_2012_GPUAcceleratedReal}, }