Comparative study on performance of 2D Image Processing Algorithm implemented in parallel and sequential models
- BALAJI VENGATESH M
- JOSEPH RAYMOND V
FPGA, CUDA, GPU, Mathematica, OpenCV
This paper deals with the Parallel Implementation of an Image Processing Algorithm and comparing the same with the sequential Implementation of the algorithm and demonstrate the Increase in speeds through Hardware rather optimizing the code for the same. We chose the Canny Edge Detection algorithm that aims at identifying points in a digital image where the image brightness changes sharply. The paper also deals with parallel Implementation on CUDALink, A Library or link provided by Wolfram Mathematica 9,which allows mathematica to use CUDA parallel computing architecture developed by NVIDIA to be used by Mathematica. We use the Mathematica’s Inbuilt Edge Detection function that runs on the CPU to compare with our parallel Implementation on the GPU. The comparison further speaks more on the quality of edges and also the running times of both the algorithm. The GPU we use here is the NVIDIA GTX 560 Ti Edition.
BALAJI VENGATESH M, JOSEPH RAYMOND V. "Comparative study on performance of 2D Image Processing Algorithm implemented in parallel and sequential models".INTERNATIONAL JOURNAL OF ENGINEERING DEVELOPMENT AND RESEARCH ISSN:2321-9939, Vol.2, Issue 1, pp.1050-1056, URL :https://rjwave.org/ijedr/papers/IJEDR1401188.pdf
Volume 2 Issue 1
Pages. 1050-1056
