An Image Denoising Algorithm based on Kuwahara Filter
Pan Guo, Xin Gong, Liying Zhang, Xuejun Li, Wenchao He, Ting Xu · 2018
In order to improve the readability of digital image, the image denoising algorithm is studied. The algorithm is performed using a combination of Kuwahara filter and non-subsampling Shearlets transform (NSST). Firstly, the image with noise is decomposed in multiple scales and in multiple directions by NSST, that is, the low frequency coefficients and high frequency coefficients are obtained, and then the threshold is processed to remove most of the noise effectively. Then, the filtered coefficient is transformed into NSST inverse transformation, and the denoising image 1 is obtained. Finally, the de-noising image 1 is filtered by the Kuwahara filter, and the final de-noising image is obtained. While de-noising, the Kuwahara filter can well retain the edge, contour and other details of the image. Through experimental simulation, this algorithm has the advantage of image denoising and can get high quality image denoising. By comparing objective evaluation indexes, this method is effective for image denoising.