Abstract:Image denoising faces significant challenges, arising from the sources of noise. Specifically, Gaussian, impulse, speckle noises, salt and pepper noise are complicated sources of noise in imaging. Convolutional neural network (CNN) has increasingly received attention in image denoising task. Due to the fast inference and good performance, discriminative learning methods have been widely studied in image denoising. However, these methods mostly learn a specific model for each noise level and require multiple models for denoising images with different noise levels. They also lack flexibility to deal with spatially variant noise, limiting their applications in practical denoising. To address these issues, we propose a deep iterative down-up CNN for image denoising, which alternates between lowering and raising the determination of the feature maps. The proposed network performed on down-sampled sub-images, achieving a good trade-off between inference speed and denoising performance. In contrast to the existing denoisers, the proposed network has several desirable properties, including (i) the ability to handle a wide range of noise levels (i.e., [0, 75]) effectively with a single network and (ii) the ability to remove spatially variant noise by specifying a non-uniform noise level map. In addition, train a network model for five different types of noises, and achieve better performance. Extensive experiments on synthetic and real noisy images are conducted to evaluate the proposed model in comparison with state-of-the-art denoisers. The results show that the proposed model is effective and efficient, making it highly attractive for practical denoising applications.

Cross-scale Color Image Restoration under High Density Salt-and-Pepper Noise

The Influence of Noise on the Restoration of Diffraction-Limited Images

Real Color Image Enhanced by Illumination - Reflectance Model and Wavelet Transformation

Real-Color Image Denoised and Enhanced Synchronously Based on Wavelet Transform

Salt and pepper noise removal method based on stationary Framelet transform with non-convex sparsity regularization

Salt-and-Pepper Noise Removal Based on Sparse Signal Processing

A Self-Supervised Hyperspectral Image Restoration Method Based on Convolutional Sparse Coding and Superpixel Segmentation

A hyperspectral image denoising method based on land cover spectral autocorrelation

Multiscale Synergism Ensemble Progressive and Contrastive Investigation for Image Restoration

Optimized the Performance of Super Resolution Images by Salt and pepper Noise Removal based on a Modified Trimmed Median Filter

Adaptive Noise Reduction Scheme for Salt and Pepper

Color image restoration using DSS-NL-mapping-based multi-noiseNet CNN model

Patch-based contour prior image denoising for salt and pepper noise

A convolutional neural networks denoising approach for salt and pepper noise

Salt-and-pepper noise removal by adaptive median filter and TV inpainting

Cross-Patch Graph Convolutional Network for Image Denoising.

Restoration Algorithm for Color Images Based on Markov Random Field

A salt and pepper noise image denoising method based on the generative classification

Speckle Image Restoration without Clean Data

An adaptive content based closer proximity pixel replacement algorithm for high density salt and pepper noise removal in images

Deep CNN for Removal of Salt and Pepper Noise