Adaptive Technique for Salt and Pepper Noise Removal through Functional Link Artificial Neural Network
Sunita Sarangi1, Suchitra Sarangi2
1Sunita Sarangi*, Asst. Prof., Department of ECE, I.T.E.R, Siksha ‘O’ Anusandhan Deemed to be University, Bhubaneswar, Odisha, India.
2Suchitra Sarangi, Asst. Prof., Department of ECE, R.I.T.E., Bhubaneswar, India.
Manuscript received on February 01, 2019. | Revised Manuscript received on February 14, 2019. | Manuscript published on December 30, 2019. | PP: 4959-4962 | Volume-9 Issue-2, December, 2019. | Retrieval Number: B3959129219/2019©BEIESP | DOI: 10.35940/ijeat.B3959.129219
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this paper, an adaptive method for removing salt and pepper noise from images is proposed. A second order difference operator is used to locate the corrupted pixels in images by comparing with a threshold, which is selected adaptively using the image properties. A functional link artificial neural network (FLANN) based method is proposed to set a threshold for each corrupted image for identification of noisy pixels using recursive zero attracting least mean square (RZALMS) as the updating algorithm. Median filter is used to eliminate noise from the detected pixel locations.
Keywords: Adaptive threshold, Median Filter, Reweighted zero Attracting LMS, Salt and Pepper noise.