A Modified Non-Linear Decision based Algorithm (MNDBA) for Noisy Images

Noise in an image varies from the ideal signal and is caused as a result of un modeled processes in the capture of the signals [1]. There are many algorithms for the removal of impulse noise. Conventional methods achieve a high degree of noise but algorithms perform poorly on mixed impulse noise. An adaptive median filter considers the positive and negative impulses for simultaneous removal but acts poorly on the strip lines, drop lines and blotches [2]. This paper proposes and investigates the performance of a new modified algorithm for noisy images, in low, mid and high noise density ranges and employs the window with smallest possible size to reduce computational time