GPU implementation of Belief Propagation method for Image Restoration using OpenCL

The image processing applications involve huge amount of computational complexity as the operations are carried out on each pixel of the image. The General Purpose computations that are data independent can run on Graphics Processing Units (GPU) to enable speedup in running time due to high level of parallelism. Compute Unified Device Architecture (CUDA) and Open Computing Language (OpenCL) programming environments are well known parallel programming languages for GPU-based Single Instruction Multiple Data (SIMD) architectures. This paper presents parallel implementation of Belief Propagation (BP) algorithm for Image Restoration on GPU using OpenCL parallel programming environment. The experimental results shows that, GPU-based implementation improves the running time of BP for image restoration when compared to sequential implmentation of BP. The best and average running time of BP algorithm on GPUs with 14 multiprocessors (48 cores) is 0.81ms and 1.46ms when tested on various benchmark images with CIF and VGA resolution.

[1]  Vladimir Kolmogorov,et al.  An Experimental Comparison of Min-Cut/Max-Flow Algorithms for Energy Minimization in Vision , 2004, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  David Kirk,et al.  NVIDIA cuda software and gpu parallel computing architecture , 2007, ISMM '07.

[3]  Steve Mann,et al.  OpenVIDIA: parallel GPU computer vision , 2005, ACM Multimedia.

[4]  P. J. Narayanan,et al.  CUDA cuts: Fast graph cuts on the GPU , 2008, 2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops.

[5]  Miao Liao,et al.  Real-time Global Stereo Matching Using Hierarchical Belief Propagation , 2006, BMVC.

[6]  Steve Mann,et al.  Using graphics devices in reverse: GPU-based Image Processing and Computer Vision , 2008, 2008 IEEE International Conference on Multimedia and Expo.

[7]  C. Kambhamettu,et al.  GPU implementation of belief propagation using CUDA for cloud tracking and reconstruction , 2008, 2008 IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS 2008).

[8]  Pedro F. Felzenszwalb,et al.  Efficient belief propagation for early vision , 2004, CVPR 2004.

[9]  Alan Brunton,et al.  Belief Propagation on the GPU for Stereo Vision , 2006, The 3rd Canadian Conference on Computer and Robot Vision (CRV'06).