Performance Analysis of Denoising Algorithm Based on Adaptive Median Filter Under Unsystematic Intensity Impulse and Salt and Pepper Noise

摘要

Since 1995, the Adaptive median filter (AMF) has been desired for denoising salt&pepper noises and this AMF filter has successfully implemented due to its powerful performance however this AMF is not desired for denoising random-valued impulse noise. Therefore, this paper comprehensively evaluates the AMF filter performance and its limitation when this AMF filter is implemented under both unsystematic intensity impulse and salt&pepper noise. In this experimental section, six noisy images (Lena, Mobile, Pepper, Pentagon, Girl, Resolution) under both salt&pepper noise and random-valued impulse noise are used to evaluate the maximum AMF filter performance in the detecting accuracy and reconstructed image quality (PSNR) perspective.