INVESTIGATING THE MEDIAN FILTER OPERATION ON CPU AND GPU

admin August 22, 2020

Authors:

Iyad Katib,

DOI NO:

https://doi.org/10.26782/jmcms.2020.08.00052

Keywords:

CUDA,GPU,Histogram Approach,Median Filter,OpenMP ,

Abstract

The Median Filter (MF) is one of the problems that need massive computational resources to perform its operation in a moderate time.  The MF can be implemented on traditional CPUs and GPUs.  Investigating the performance in terms of processing time of the MF on different architectures can provide the researchers with wider vision to optimally select the computational resources that best fit the required time needed to remove salt and pepper noise.  This paper shows the impact of different parameters affecting the MF processing time.  Resolution of the frame, frame rate per second, and the MF r value are investigated in order to decide both the preferred architecture and algorithm.  OpenMP has been deployed on CPUs and CUDA has been deployed on Nvidia GPGPU K20.  Experimental results show that histogram approach and K20 using CUDA are the best choice for processing 4K resolution with r > 2 and HD resolution with r > 4. For VGA resolution and r > 6, histogram approach and CPU using OpenMP are the best choice.  The paper provides a way to select the architecture-algorithm pair suitable for implementing the MF

Refference:

I. C. M. Wu and Y. C. Chiang, “Insertion Sort Circuit Design Applied on the Median Filter,” in 2018 IEEE International Conference on Consumer Electronics-Taiwan, ICCE-TW 2018, 2018.

II. D. S. Richards, “VLSI Median Filters,” IEEE Trans. Acoust., 1990.

III. G. Gupta, “Algorithm for Image Processing Using Improved Median Filter and Comparison of Mean, Median and Improved Median Filter,” Int. J. Soft Comput., 2011.

IV. H. M.Faheem and B. König-Ries, “A New Scheduling Strategy for Solving the Motif Finding Problem on Heterogeneous Architectures,” Int. J. Comput. Appl., 2014.

V. K. Verma, B. Kumar Singh, and A. S. Thokec, “An enhancement in adaptive median filter for edge preservation,” in Procedia Computer Science, 2015.

VI. L. Hayat, M. Fleury, and A. F. Clark, “Two-dimensional median filter algorithm for parallel reconfigurable computers,” IEE Proc. Vision, Image Signal Process., 1995.

VII. M. Fayez, H. M. Faheem, I. Katib, and N. R. Aljohani, “Real-time image scanning framework using GPGPU – Face detection case study,” in Proceedings of the 2016 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2016, 2016.

VIII. M. Vega-Rodríguez and J. Sánchez-Pérez, “An FPGA-based implementation for median filter meeting the real-time requirements of automated visual inspection systems,” Proc. 10th IEEE Mediterr. Conf. Control Autom. (MED ’02), 2002.

IX. N. A. Sabour, H. M. Faheem, and M. E. Khalifa, “Multi-agent based framework for target tracking using a real time vision system,” in 2008 International Conference on Computer Engineering and Systems, ICCES 2008, 2008.

X. O. Green, “Efficient scalable median filtering using histogram-based operations,” IEEE Trans. Image Process., 2018.

XI. R. Medhat, H. M. Faheem, and M. E. Khaleefa, “Efficient parallel architecture of median filter,” in Proceedings of the 9th IASTED International Conference on Parallel and Distributed Computing and Networks, PDCN 2010, 2010.

XII. S. Perreault and P. Hébert, “Median filtering in constant time,” IEEE Trans. Image Process., 2007.

XIII. Y. He, P. Liu, Z. Wang, Z. Hu, and Y. Yang, “Filter pruning via geometric median for deep convolutional neural networks acceleration,” in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2019.

View Download