Image Quality Improvement of Surveillance Camera Images by Noise Removal Using Learning Method
Akira Kuchida, Tomio Goto · 2021 IEEE 10th Global Conference on Consumer Electronics (GCCE) · 2021
In this paper, we propose an image denoising method that combines the Total Variation (TV) regularization method with a learning method of Denoising Convolutional Neural Network (DnCNN). In this method, a texture component loss, which is one of the problems of learning methods, is addressed by performing image decompositions using the TV regularization method to denoise the image without losing the texture components. As a result, the subjective evaluation was improved over the conventional method in terms of denoising performance and texture preservation.