A GPU Accelerated Algorithm for Compressive Sensing Based Image Super-Resolution

Author

Wu, Xifei ; Xiang, Hui ; Lu, Peng

Author_Institution

Sch. of Comput. Sci. & Technol., Shandong Univ., Jinan, China

fYear

2011

fDate

15-16 May 2011

Firstpage

198

Lastpage

202

Abstract

This paper presents a parallel algorithm designed for Super-resolution Image Reconstruction based on Compressive sensing in the ATI Stream platform. In the accelerating process, we select part of the serial program as the objects to be sped up according to the execution time of each stage, set appropriate parallel granularity to make full use of GPU´s computational horsepower, and make a rational use of different kinds of memory space in GPU. At last, the result of the parallel algorithm is shown and analyzed. Compared to the serial algorithm, parallel algorithm has significantly accelerated results.

Keywords

coprocessors; image reconstruction; image resolution; signal representation; ATI Stream platform; GPU accelerated algorithm; compressive sensing; image reconstruction; image super-resolution; parallel algorithm; serial program; Acceleration; Compressed sensing; Graphics processing unit; Image reconstruction; Image resolution; Kernel; Signal resolution; Compressive Sensing; GPU; Image Super-resolution; parallel computing;

fLanguage

English

Publisher

ieee

Conference_Titel

Digital Media and Digital Content Management (DMDCM), 2011 Workshop on

Conference_Location

Hangzhou

Print_ISBN

978-1-4577-0271-6

Electronic_ISBN

978-0-7695-4413-7

Type

conf

DOI

10.1109/DMDCM.2011.10

Filename

5959700