Abstract:Image denoising remains a classical yet still challenging problem, because the noise can destroy details and cause severe information loss. In recent years, various well-designed CNN-based methods have been extensively applied in image denoising because of the strong learning ability. However, most of them share an unidirectional procedure, which directly learns a mapping from noisy input to a clean image without focusing on the over-smoothed state of the denoising process, limiting the richness of extracted features. Different from previous works, we propose a blurred image feature guided CNN (BFCNN) network that implements a novel blurring-adjusting strategy to address the complex denoising problem via two stages. In stage 1, we build a blurring module (BM) to capture over-smoothed features from noisy observations and generate the blurred image restoration, which is a less informative version of the clean image. Furthermore, a multi-level concatenating module (CM) and an adjusting module (AM) are then designed to recover more detailed information in stage 2. These two modules are jointly designed to restore a properly-smoothed image from the over-smoothed blurred image and the given under-smoothed noisy image. Comparing to the traditional denoising process, the proposed blurring-adjusting strategy produces a precise denoised image more efficiently by converting the unidirectional denoising process into a bidirectional denoising process. To our knowledge, this is the first study that utilizes the over-smoothed image to address the denoising problem. Extensive experiments demonstrate the superiority of our BFCNN with more accurate reconstruction quality and achieve competitive quantitative results among current CNN-based methods. This research will release the code upon acceptance.

Deep CNN denoiser prior for blurred images restoration with multiplicative noise

Ground-to-air Aircraft Infrared Image Deblurring Based on Imaging Degradation Simulation

Deep CNN Denoiser prior for multiplicative noise removal

Nonblind Image Deblurring Via Deep Learning in Complex Field

A Prior-Guided Deep Network for Real Image Denoising and Its Applications

Deep Denoiser Prior Based Deep Analytic Network for Lensless Image Restoration.

Learning spectral-spatial prior via 3ddncnn for hyperspectral image deconvolution

Image restoration for synthetic aperture systems with a non-blind deconvolution algorithm via a deep convolutional neural network

Learning Deep CNN Denoiser Prior for Image Restoration

Bidirectional Image Denoising with Blurred Image Feature

Learning Blind Denoising Network for Noisy Image Deblurring.

A CNN-Based Self-Supervised Synthetic Aperture Radar Image Denoising Approach

Hyperspectral Image Denoising Based on Deep and Total Variation Priors

Plug-and-Play Image Restoration with Deep Denoiser Prior

Attention-guided CNN for image denoising

Denoising Prior Driven Deep Neural Network for Image Restoration

Toward convolutional blind denoising of real photographs

Blind Image Deblurring Via Deep Discriminative Priors

Deep learning informed diffusion equation model for image denoising