GPU Implementation of a Clustering Based Image Registration

Authors:
Seung-Hun Yoo;Yun-Seok Lee;Sung-Up Jo;Chang-Sung Jeong
Affiliations:
Department of Electronics and Computer Engineering, Korea University, Seoul, Korea;Department of Electronics and Computer Engineering, Korea University, Seoul, Korea;Department of Electronics and Computer Engineering, Korea University, Seoul, Korea;Department of Electronics and Computer Engineering, Korea University, Seoul, Korea
Venue:
ICIC '08 Proceedings of the 4th international conference on Intelligent Computing: Advanced Intelligent Computing Theories and Applications - with Aspects of Theoretical and Methodological Issues
Year:
2008

Citing 7
Cited 0

A Computational Approach to Edge Detection

IEEE Transactions on Pattern Analysis and Machine Intelligence
The Cg Tutorial: The Definitive Guide to Programmable Real-Time Graphics

The Cg Tutorial: The Definitive Guide to Programmable Real-Time Graphics
Scout: A Hardware-Accelerated System for Quantitatively Driven Visualization and Analysis

VIS '04 Proceedings of the conference on Visualization '04
2-D and 3-D Image Registration: for Medical, Remote Sensing, and Industrial Applications

2-D and 3-D Image Registration: for Medical, Remote Sensing, and Industrial Applications
GPGPU: general purpose computation on graphics hardware

ACM SIGGRAPH 2004 Course Notes
Automatic Aerial Image RegistrationWithout Correspondence

ICVS '06 Proceedings of the Fourth IEEE International Conference on Computer Vision Systems
Discrete Wavelet Transform on Consumer-Level Graphics Hardware

IEEE Transactions on Multimedia

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents GPU(Graphics Processing Unit) implementation of a clustering based image registration method. Since the image registration is an important process in image analysis tasks such as image restoration and image fusion, fast image registration can improve the overall application execution speed. Recently, the commodity GPU is being used in not only 3D graphics rendering but also in general-purpose computation due to an increase in the good price/performance ratio and hardware programmability as well as the huge computing power and speed of the GPU. We implemented clustering-based image registration method on GPU using only transformation of texture coordinations in vertex program and re-sampling in fragment program. Finally, GPU-based image registration speed up nearly 50 percent compared with CPU.