A New Self-Adaptive Weighted Filter for Removing Noise in Infrared Images

Abstract

Aimed at the infrared images with mixed noise, this paper proposes a new adaptive weighted mixed filtering algorithm. The new algorithm first judges the type of the noise according to the difference values of pixel's neighborhood region. Then removes the gauss noise by self-adaptive weighted mean filter and removes the impulse noise by self-adaptive weighted median filter. The result of computer simulation on infrared images indicates that the new algorithm is superior to traditional mean filtering algorithm and traditional median filtering algorithm. The new algorithm can not only remove mixed noise effectively, but also reserve image details.