Thresholding-based image denoising using twodimensional S-transform

S-transform is an effective time-frequency analysis technique which can provide simultaneous time and frequency distribution information similar to the wavelet transform. It is a multi-resolution analysis method applicable to image denoising and is expected to become a powerful processing tool in this field. We introduce the ideas of wavelet transform denoising into Stransform and propose thresholding-based image denoising using two-dimensional S-transform. Simulation experimental results illustrate the favorable application performance of the method, while its successful implementation in myocardial contrast echocardiography image denoising demonstrates its clinical value. © 2012 IEEE.