Image Denoising Using Nonlocal Regularized Deep Image Prior

Deep neural networks have shown great potential in various low-level vision tasks, leading to several state-of-the-art image denoising techniques. Training a deep neural network in a supervised fashion usually requires the collection of a great number of examples and the consumption of a significant amount of time. However, the collection of training samples is very difficult for some application scenarios, such as the full-sampled data of magnetic resonance imaging and the data of satellite remote sensing imaging. In this paper, we overcome the problem of a lack of training data by using an unsupervised deep-learning-based method. Specifically, we propose a deep-learning-based method based on the deep image prior (DIP) method, which only requires a noisy image as training data, without any clean data. It infers the natural images with random inputs and the corrupted observation with the help of performing correction via a convolutional network. We improve the original DIP method as follows: Firstly, the original optimization objective function is modified by adding nonlocal regularizers, consisting of a spatial filter and a frequency domain filter, to promote the gradient sparsity of the solution. Secondly, we solve the optimization problem with the alternating direction method of multipliers (ADMM) framework, resulting in two separate optimization problems, including a symmetric U-Net training step and a plug-and-play proximal denoising step. As such, the proposed method exploits the powerful denoising ability of both deep neural networks and nonlocal regularizations. Experiments validate the effectiveness of leveraging a combination of DIP and nonlocal regularizers, and demonstrate the superior performance of the proposed method both quantitatively and visually compared with the original DIP method.

Efficient Fundus Image Gradeability Approach Based on Deep Reconstruction-Classification Network

Quality of retinal image is vital for screening of ailments pertaining to eye such as glaucoma, diabetic retinopathy (DR) and age related macular degeneration. Therefore, assessing quality of retinal image prior to any kind of diagnosis has assumed significance in Computer Aided Desgin (CAD) applications. The rationale behind this is that reliability of retinal image is to be guaranteed to have dependable diagnosis. In this paper, we propose a novel retinal fundus image quality assessment (RIQA) method based on autoencoder network to assess retinal images if the image is acceptable for screening or not. The autoencoder network architecture is well suited to precisely to properly represent the key features of the image quality, especially when the network can correctly reconstruct the input image. The proposed model consists of encoder and decoder successive networks. The encoder will be used for representing the features of the input image. In turn , the decoder will be used for reconstruct the input image. The features get from encoder network will then be fed to a classifier in order to classify the quality of retinal image to two classes: gradable or ungradable. The experimental results revealed more useful assessment and the proposed deep model provides a superior performance for RIQA. Thus, our model can serve real-world Clinical Decision Support Systems in the healthcare domain.

Hyperspectral Image Mixed Noise Removal Using Subspace Representation and Deep CNN Image Prior

The ever-increasing spectral resolution of hyperspectral images (HSIs) is often obtained at the cost of a decrease in the signal-to-noise ratio (SNR) of the measurements. The decreased SNR reduces the reliability of measured features or information extracted from HSIs, thus calling for effective denoising techniques. This work aims to estimate clean HSIs from observations corrupted by mixed noise (containing Gaussian noise, impulse noise, and dead-lines/stripes) by exploiting two main characteristics of hyperspectral data, namely low-rankness in the spectral domain and high correlation in the spatial domain. We take advantage of the spectral low-rankness of HSIs by representing spectral vectors in an orthogonal subspace, which is learned from observed images by a new method. Subspace representation coefficients of HSIs are learned by solving an optimization problem plugged with an image prior extracted from a neural denoising network. The proposed method is evaluated on simulated and real HSIs. An exhaustive array of experiments and comparisons with state-of-the-art denoisers were carried out.