Truncated total variation image denoising model based on fractional B-spline wavelet

Abstract

In order to suppress the noise and keep the clear structure, edges and details of the image under noisy images, a truncated total variation (truncated TV) image denoising model based on the fractional wavelet domain is proposed. The fractional B-spline wavelet can well describe the features of fractional singularity such as fine textures in the image. The truncated TV model is based on the TV model, which solves the shortcomings of the TV model’s large penalty amplitude and reduced image contrast. The model first performs the fractional B-spline wavelet transform on the image, applies truncated TV to the fractional wavelet domain to process the image, and then performs the inverse transform of the fractional B-spline wavelet transform to obtain the denoised image. The experimental results of image noise reduction show that compared with some existing image noise reduction methods, the proposed noise reduction model can effectively suppress noise and maintain the structure and details of the image. The application of the model to other fields in the future remains to be studied.