Noise Reduction, Enhancement and Classification for Sonar Images

• DOI: 10.24996/ijs.2021.62.11(si).25
• Publication Date: 2021-12-30T10:51:54Z
• Authors: Noor H. Resham; Heba Kh. Abbas; Haidar J. Mohamad; Anwar H. Al-Saleh
• Keywords: Speckle noise; Salt-and-pepper noise; Image noise; Gradient noise; Multiplicative noise; Value noise; Wiener filter

Ultrasound imaging has some problems with image properties output. These affects the specialist decision. noise type is speckle which a grainy pattern depending on signal. There are two parts of this study. The first part enhancing images adaptive Weiner, Lee, Gamma and Frost filters 3x3, 5x5, 7x7 sliding windows. evaluated process was achieved using signal to ratio (SNR), peak (PSNR), mean square error (MSE), maximum difference (MD) criteria. second consists simulating in standard (Lina image) by adding different percentage from 0.01 0.06. supervised classification based minimum distance method used evaluate results selecting four blocks located at places image. Speckle added 0.06 calculate coherent within concluded slope deviation for each noise. showed that additive increased slide window size, while multiplicative did not change nor increasing ratio. Wiener filter best