Image Noise Removal Method Based on Thresholding and Regularization Techniques

In this article, a salt and pepper noise (SPN) removal method is proposed based on thresholding regularization techniques. The utilizes the ability to remove from an image denoising model Total Variation (TV) characteristics of SPN. First, technique characteristic SPN detect noisy pixels. Second, modified TV regularization-based applied restore above addition, numerical implementation Nesterov optimal also provided. Five test cases with various levels for large natural dataset were studied in experiments. peak signal-to-noise ratio structural similarity metrics employed assess quality after denoising. experimental results indicated that removes remarkably outperforms state-of-the-art methods