ICTACT Journal on Image and Video Processing

K. B. Sangavi ¹, Nirmala Madian ¹, E. Mythili ¹, S. Jayasri ²

Affiliations
1 Department of Computer Science and Engineering, Sri Shakthi Institute of Engineering and Technology, IN
2 Nano Electronics and Integration Division, IRRD Automatons, IN

Abstract

This paper suggests an effective algorithm based on the average variance from the digital images to eliminate salt and pepper noise. The proposed algorithm chooses the 33 window to unsymmetrically decorate the corrupted pixel and to swap the corrupted pixel for the median of the other pixel. In the chosen pane, on the other side, the screen width will be decreased by two when the whole pixel comprises 0 and 255 and the same process will be replicated. When noisy pixel values cannot even be achieved in a 7 × 7 frame, then the main pixel will be substituted with a small statistical variance. Experimental results show that the algorithm suggested continuously operates to reduce noise from salt and pepper. The unbiased, analytical analysis of the proposed algorithm shows that the proposed algorithm beats the current state-of-the-art algorithm for noise reduction, such as SMF, AMF, DBA and MDBUTMF.

Keywords

Adaptive Median Filter, Decision Based Algorithm, Mean Deviation and Standard Median Filter.

Full Text