2D-VMD Based Speckle Reduction in Ultrasound Images using JS distance and Wavelet Thresholding

Abstract

This paper proposes a technique called 2D-VMD Based Speckle Reduction in Ultrasound Images Using Jensen Shannon (JS) Distance and Wavelet Thresholding. First, the image is broken up into a set of functions known as intrinsic mode functions (IMFs). The JS distance between the probability density function (PDF) is the measure to tell the difference between IMFs that are dominated by signal and IMFs that are dominated by noise. The IMFs that depend on noise are then cleaned up using wavelet threshold denoising method. The signal-dependent IMFs and the denoised IMFs are used to rebuild the image and get the denoised image. The proposed technique is tested on simulated and real ultrasound images with different levels of noise. In terms of EKI, CC, PSNR, SSIM, MVR and ENL, the experiments confirms that the proposed algorithm is better than other methods that are already in use.