An efficient denoising of impulse noise from MRI using adaptive switching modified decision based unsymmetric trimmed median filter

Abstract

Noise reduction in digital images is the eminent preprocessing scheme in major image processing activities which includes medical image segmentation. A well designed noise reduction scheme for highly corrupted medical images is needed for medical diagnosing schemes with an immediate attention. This paper proposes a novel Adaptive Switching Modified Decision Based Unsymmetric Trimmed Median Filter (ASMDBUTMF) for noise reduction in gray scale MR Images which are affected by salt and pepper noise. The Proposed method is the extension work of the existing Modified Decision Based Unsymmetric Trimmed Median Filter (MDBUTMF) method. In this paper noisy pixels are detected using the occurrence of intensity values 0’s and 255’s. The noisy pixels are converted into noise-free pixels using trimmed median value s. The MDBUTMF Filter uses 3x3 size windows only while the proposed method uses window sizes based on Adaptive switching method to retain more details for noise reduction. The proposed ASMDBUTMF algorithm is compared with existing methods, namely Median Filter (MF), Rank Order Median Filter (ROMF), Noise Adaptive Fuzzy Switching Median Filter (NAFSMF), Modified Decision Based Unsymmetric Trimmed Median Filter (MDBUTMF) and Geometric Multiscale Ridgelet Support Vector Transform Filter (GMRSVTF) simultaneously. The new ASMDBUTMF Filter supports noise reduction for highly corrupted images with better peak signal-to-noise Ratio (PSNR) and Image Enhancement Factor (IEF).