Performance Improvement of 1-D Gaussian Filter using SIMD Technology
Paper ID : 1072-MVIP2020
maryam moradifar *1, Asadollah Shahbahrami2
2دانشگاه گیلان
Denoising is an important process before applying other post-processing techniques on medical images. To obtain better quality images many denoising approaches have been introduced. Gaussian filter is a spatial domain filter, which is proper to deblur and remove noise from images. Since the Gaussian filter modifies the input signal by convolution with a Gaussian function it is a computation intensive algorithm. To enhance the performance of the algorithm, it is better to perform two 1-D convolution operations instead of one 2-D convolution operation and then parallelize it. In this paper in order to increase the performance of 1-D Gaussian filter, we exploit both DLP and TLP using parallel programming models such as IPM, CAVs and OpenMP. The experimental results were shown that the performance of our implementations is much higher than other approaches.
Gaussian filter, 1-D convolution, Single Instruction Multiple Data, Data-level Parallelism, Thread-level Parallelism
Status : Paper Accepted (Oral Presentation)