A comparative analyses of training of separable and non-separable image filters with genetic algorithms
Abstract
Separable image filter is a subclass of convolutional image filters that are used widely. The coefficients of these types of image filters can be determined with training images using heuristic approaches as well as analytical methods. In this study, comparative analyses were realized for separable and non-separable image filters that were trained using genetic algorithms. The results for training durations and performance analyses are presented comparatively for various size of kernels. According to the results, the training durations of the separable image filter is shorter due to smaller number of coefficients and hence smaller number of multiplication and addition operations. On the other hand, when compared in terms of quality, non-separable filter shows better results.
Keywords
References
- [1] M. Y. Çelikdemir, “Beton Yapılarda Görüntü Filtreleme Tekniklerinin Uygulanması,” 2014, pp. 308–310.
- [2] D. Akgün, “A practical parallel implementation for TDLMS image filter on multi-core processor,” J. Real-Time Image Process., Jan. 2014.
- [3] D. Akgün, “TDLMS Filtresinin Görüntü İyleştirme İçin Başarım Analizi,” NWSA Eng. Sci., 2011.
- [4] R. C. Gonzalez, R. E. Woods, and S. L. Eddins, “Digital Image Processing Using Matlab - Gonzalez Woods & Eddins.pdf,” Education, vol. 624, no. 2. p. 609, 2004.
- [5] F. Talbi, F. Alim, S. Seddiki, and I. Mezzah, “Separable convolution gaussian smoothing filters on a xilinx FPGA platform,” (INTECH), 2015 Fifth …, 2015.
- [6] P. Narendra, “A separable median filter for image noise smoothing,” Mach. Intell. IEEE Trans., 1981.
- [7] M. Siekmann, S. Bosse, and H. Schwarz, “Separable Wiener filter based adaptive in-loop filter for video coding,” Pict. Coding, 2010.
- [8] Q. Ma and C. Cowan, “Genetic algorithms applied to the adaptation of IIR filters,” Signal Processing, 1996.
Details
Primary Language
English
Subjects
Computer Software
Journal Section
Research Article
Publication Date
August 1, 2017
Submission Date
June 29, 2016
Acceptance Date
April 20, 2017
Published in Issue
Year 2017 Volume: 21 Number: 4