Explore the vast range of titles available.

Self-similarity refers to the image prior widely used in image restoration algorithms that small but similar patterns tend to occur at different locations and scales. However, recent advanced deep convolutional neural network-based methods for image restoration do not take full advantage of self-similarities by relying on self-attention neural modules that only process information at the same scale. To solve this problem, we present a novel Pyramid Attention module for image restoration, which captures long-range feature correspondences from a multi-scale feature pyramid. Inspired by the fact that corruptions, such as noise or compression artifacts, drop drastically at coarser image scales, our attention module is designed to be able to borrow clean signals from their “clean” correspondences at the coarser levels. The proposed pyramid attention module is a generic building block that can be flexibly integrated into various neural architectures. Its effectiveness is validated through extensive experiments on multiple image restoration tasks: image denoising, demosaicing, compression artifact reduction, and super resolution. Without any bells and whistles, our PANet (pyramid attention module with simple network backbones) can produce state-of-the-art results with superior accuracy and visual quality. Our code is available at https://github.com/SHI-Labs/Pyramid-Attention-Networks

Snapshot polarization imaging has gained interest in the last few decades. Recent research and technology achievements defined the polarization Filter Array (PFA). It is dedicated to division-of-focal plane polarimeters, which permits to analyze the direction of light electric field oscillation. Its filters form a mosaicked pattern, in which each pixel only senses a fraction of the total polarization states, so the other missing polarization states have to be interpolated. As for Color or Spectral Filter Arrays (CFA or SFA), several dedicated demosaicking methods exist in the PFA literature. Such methods are mainly based on spatial correlation disregarding inter-channel correlation. We show that polarization channels are strongly correlated in images. We therefore propose to extend some demosaicking methods from CFA/SFA to PFA, and compare them with those that are PFA-oriented. Objective and subjective analysis show that the pseudo panchromatic image difference method provides the best results and can be used as benchmark for PFA demosaicking.

The simultaneous acquisition of multi-spectral images on a single sensor can be efficiently performed by single shot capture using a mutli-spectral filter array. This paper focused on the demosaicing of color and near-infrared bands and relied on a convolutional neural network (CNN). To train the deep learning model robustly and accurately, it is necessary to provide enough training data, with sufficient variability. We focused on the design of an efficient training procedure by discovering an optimal training dataset. We propose two data selection strategies, motivated by slightly different concepts. The general term that will be used for the proposed models trained using data selection is data selection-based multi-spectral demosaicing (DSMD). The first idea is clustering-based data selection (DSMD-C), with the goal to discover a representative subset with a high variance so as to train a robust model. The second is an adaptive-based data selection (DSMD-A), a self-guided approach that selects new data based on the current model accuracy. We performed a controlled experimental evaluation of the proposed training strategies and the results show that a careful selection of data does benefit the speed and accuracy of training. We are still able to achieve high reconstruction accuracy with a lightweight model.

The methods employed for image demosaicing and denoising play a pivotal role in image acquisition and restoration, and have been extensively studied over the past few decades. Traditionally, these tasks are performed sequentially, with demosaicing followed by denoising, or vice versa, treating each process independently. While this approach can enhance image quality, it often leads to issues such as color inaccuracies and information loss, as the outcome of the first task influences the second. Consequently, the integration of joint demosaicing and denoising (JDD) has become a focal point in recent research. Deep convolutional neural networks have shown promising results in addressing JDD challenges. This study introduces an end-to-end network, termed the Frequency-domain Features learning Network (FFNet), designed to tackle the JDD problem. Unlike conventional methods that focus on spatial domain features, FFNet utilizes frequency-domain (FD) characteristics to capture both global and local image details. Based on the vision Transformer architecture, FFNet consists of two key components: a global Fourier block (GFB), which uses global attention to determine the weights of FD parameters, and an MLP-based local Fourier block (LFB), which improves local feature extraction. These blocks are integrated with a channel attention mechanism to form the frequency-domain attention block (FAB), the core element of FFNet. Extensive experimental results on benchmark datasets demonstrate that FFNet achieves superior performance in terms of both quantitative metrics (PSNR/SSIM) and visual quality compared to existing state-of-the-art JDD methods. Furthermore, we provide a comprehensive analysis of its computational efficiency, including parameter count, FLOPs, and inference time, showing a competitive trade-off between performance and complexity.

This paper proposes a method for demosaicing raw images captured by multispectral cameras. The proposed method estimates a pseudo-panchromatic image (PPI) via an iterative-linear-regression model and utilizes the estimated PPI for multispectral demosaicing. The PPI is estimated through horizontal and vertical guided filtering, with the subsampled multispectral-filter-array-(MSFA) image and low-pass-filtered MSFA as the guide image and filtering input, respectively. The number of iterations is automatically determined according to a predetermined criterion. Spectral differences between the estimated PPI and MSFA are calculated for each channel, and each spectral difference is interpolated using directional interpolation. The weights are calculated from the estimated PPI, and each interpolated spectral difference is combined using the weighted sum. The experimental results indicate that the proposed method outperforms the State-of-the-Art methods with regard to spatial and spectral fidelity for both synthetic and real-world images.

Polarization imaging has achieved a wide range of applications in military and civilian fields such as camouflage detection and autonomous driving. However, when the imaging environment involves a low-light condition, the number of photons is low and the photon transmittance of the conventional Division-of-Focal-Plane (DoFP) structure is small. Therefore, the traditional demosaicing methods are often used to deal with the serious noise and distortion generated by polarization demosaicing in low-light environment. Based on the aforementioned issues, this paper proposes a model called Low-Light Sparse Polarization Demosaicing Network (LLSPD-Net) for simulating a sparse polarization sensor acquisition of polarization images in low-light environments. The model consists of two parts: an intensity image enhancement network and a Stokes vector complementation network. In this work, the intensity image enhancement network is used to enhance low-light images and obtain high-quality RGB images, while the Stokes vector is used to complement the network. We discard the traditional idea of polarization intensity image interpolation and instead design a polarization demosaicing method with Stokes vector complementation. By using the enhanced intensity image as a guide, the completion of the Stokes vector is achieved. In addition, to train our network, we collected a dataset of paired color polarization images that includes both low-light and regular-light conditions. A comparison with state-of-the-art methods on both self-constructed and publicly available datasets reveals that our model outperforms traditional low-light image enhancement demosaicing methods in both qualitative and quantitative experiments.

Multi-spectral imaging technologies have made great progress in the past few decades. The development of snapshot cameras equipped with a specific multi-spectral filter array (MSFA) allow dynamic scenes to be captured on a miniaturized platform across multiple spectral bands, opening up extensive applications in quantitative and visualized analysis. However, a snapshot camera based on MSFA captures a single band per pixel; thus, the other spectral band components of pixels are all missed. The raw images, which are captured by snapshot multi-spectral imaging systems, require a reconstruction procedure called demosaicing to estimate a fully defined multi-spectral image (MSI). With increasing spectral bands, the challenge of demosaicing becomes more difficult. Furthermore, the existing demosaicing methods will produce adverse artifacts and aliasing because of the adverse effects of spatial interpolation and the inadequacy of the number of layers in the network structure. In this paper, a novel multi-spectral demosaicing method based on a deep convolution neural network (CNN) is proposed for the reconstruction of full-resolution multi-spectral images from raw MSFA-based spectral mosaic images. The CNN is integrated with the channel attention mechanism to protect important channel features. We verify the merits of the proposed method using 5 × 5 raw mosaic images on synthetic as well as real-world data. The experimental results show that the proposed method outperforms the existing demosaicing methods in terms of spatial details and spectral fidelity.

A joint demosaicing and denoising task refers to the task of simultaneously reconstructing and denoising a color image from a patterned image obtained by a monochrome image sensor with a color filter array. Recently, inspired by the success of deep learning in many image processing tasks, there has been research to apply convolutional neural networks (CNNs) to the task of joint demosaicing and denoising. However, such CNNs need many training data to be trained, and work well only for patterned images which have the same amount of noise they have been trained on. In this paper, we propose a variational deep image prior network for joint demosaicing and denoising which can be trained on a single patterned image and works for patterned images with different levels of noise. We also propose a new RGB color filter array (CFA) which works better with the proposed network than the conventional Bayer CFA. Mathematical justifications of why the variational deep image prior network suits the task of joint demosaicing and denoising are also given, and experimental results verify the performance of the proposed method.

Residual interpolations are effective methods to reduce the instantaneous field-of-view error of division of focal plane (DoFP) polarimeters. However, their guide-image selection strategies are improper, and do not consider the DoFP polarimeters’ spatial sampling modes. Thus, we propose a residual interpolation method with a new guide-image selection strategy based on the spatial layout of the pixeled polarizer array to improve the sampling rate of the guide image. The interpolation performance is also improved by the proposed pixel-by-pixel, adaptive iterative process and the weighted average fusion of the results of the minimized residual and minimized Laplacian energy guide filters. Visual and objective evaluations demonstrate the proposed method’s superiority to the existing state-of-the-art methods. The proposed method proves that considering the spatial layout of the pixeled polarizer array on the physical level is vital to improving the performance of interpolation methods for DoFP polarimeters.

Single-chip multispectral imaging sensors with vertically stacked photodiodes and pixelated spectral filters enable advanced, real-time visualization for image-guided cancer surgery. However, their design inherently reduces spatial resolution. We present a convolutional neural network (CNN)-transformer demosaicing algorithm, validated on both clinical and preclinical datasets that effectively doubles spatial resolution and improves image quality-substantially enhancing intraoperative cancer visualization. We present a CNN-transformer-based demosaicing approach specifically optimized for reconstructing high-resolution color and NIR images acquired by a hexachromatic imaging sensor. A hybrid CNN-transformer demosaicing model was developed and trained on color-image datasets, then rigorously evaluated on color and NIR images to demonstrate superior reconstruction quality compared with conventional bilinear interpolation and residual CNN methods. Our CNN-transformer demosaicing method achieves an average mean squared error (MSE) reduction of for color images and 76% for NIR images and improves structural dissimilarity by roughly 72% and 79%, respectively, compared with state-of-the-art CNN-based demosaicing algorithms in preclinical datasets. In clinical datasets, our approach similarly demonstrates significant reductions in MSE and structural dissimilarity, substantially outperforming existing CNN-based methods, particularly in reconstructing high-frequency image details. We demonstrate improvements in spatial resolution and image fidelity for color and NIR images obtained from hexachromatic imaging sensors, achieved by integrating convolutional neural networks with transformer architectures. Given recent advances in GPU computing, our CNN-transformer approach offers a practical, real-time solution for enhanced multispectral imaging during cancer surgery.