Enhanced Ultrasound Speckle Reduction technique Using SLIC Segmentation and Method Noise Compensation

Abstract

In this paper, an enhanced ultrasound speckle reduction technique using Simple linear iterative clustering (SLIC) segmentation and method noise compensation technique has been proposed. The noisy image is first sparsely represented in its low frequency components with the SLIC superpixel clustering and reconstruction. Then the high frequency method noise is calculated by the difference of the noisy image and the low frequency reconstructed image. The details in the method image are then utilized to enhance the low frequency reconstructed image. The performance of the proposed technique is tested on synthetic ultrasound images with various noise components of different variance. The experiments show promising results of the proposed algorithm in comparison to other existing methods in terms of PSNR, SSIM, EKI and CC.