Segment Anything Models (SAMs), as vision foundation models, have demonstrated remarkable performance across various image analysis tasks. Despite their strong generalization capabilities, SAMs encounter challenges in fine-grained detail segmentation for high-resolution class-independent segmentation (HRCS), due to the limitations in the direct processing of high-resolution inputs and low-resolution mask predictions, and the reliance on accurate manual prompts. To address these limitations, we propose MGD-SAM2, which integrates SAM2 with multi-view feature interaction between a global image and local patches to achieve precise segmentation. MGD-SAM2 incorporates the pre-trained SAM2 with four novel modules: the Multi-view Perception Adapter (MPAdapter), the Multi-view Complementary Enhancement Module (MCEM), the Hierarchical Multi-view Interaction Module (HMIM), and the Detail Refinement Module (DRM). Specifically, we first introduce MPAdapter to adapt the SAM2 encoder for enhanced local-global perception required by HRCS images. Then, MCEM and HMIM are proposed to further enrich the local texture and global semantics by aggregating multi-view features within and across multi-scales. Finally, DRM is designed to generate gradually restored high-resolution mask predictions by jointly leveraging mask features and original images, compensating for the detail loss caused by directly upsampling low-resolution prediction maps. Experimental results demonstrate the superior performance and strong generalization of our model on multiple high-resolution and normal-resolution datasets, while requiring only 8.8% trainable parameters. Code is available at https://github.com/sevenshr/MGD-SAM2.

Light propagation underwater is susceptible to wavelength attenuation and scattering, leading to degradation plagued by color distortion, contrast degradation, and reduced visibility in underwater imaging. To handle the degradations, the paper proposes an underwater image enhancement method via advantage feature weighted fusion, called AFWF. Specifically, we propose a three-channel contrast enhancement strategy that effectively reduces the color distortion of a raw input image via a three-channel adaptive color compensation strategy. Meanwhile, we employ a fast exposure fusion to integrate the image sequences obtained from the multi-scale gamma correction and adaptive contrast enhancement strategies to improve the global contrast of the above-mentioned image. Subsequently, a single-channel contrast enhancement is proposed to improve the local contrast and edge detail information by enhancing the multi-level details of the raw image. Finally, we adopt the advantage feature weighted fusion strategy to analyze and selectively fuse the advantage feature of different enhanced images layer by layer to reconstruct a high-quality result. Extensive experimental verification results highlight that our AFWF method is superior to the state-of-the-art (SOTA) methods in improving raw underwater images’ color, contrast, and detail. The code is publicly available at: https://www.researchgate.net/publication/393021384_2025-AFWF.

Underwater images often suffer from visual degradation due to varying light absorption at different wavelengths and scattering from suspended particles. To tackle these issues, we present an intelligent optimized multi-exposure image fusion method called IMIF. Specifically, we propose an adaptive color transfer strategy that employs a colorless reference image to correct the color distortion issue by transferring the mean and standard deviation of the reference image to adjust a color-balanced image. Subsequently, we introduce a particle swarm optimization algorithm that intelligently selects the optimal set of exposure image sequences by employing information entropy and edge intensity of the image as fitness metrics. Meanwhile, we leverage a guided filtering strategy to decompose the exposure image sequences into basic and detailed layers, taking into account the exposure characteristics of each layer to generate corresponding weight maps. Finally, we employ a multi-exposure fusion strategy to adaptively fuse the exposed image sequences with weight maps, producing an enhanced result. Extensive experiments conducted on three datasets demonstrate that our IMIF method outperforms state-of-the-art (SOTA) methods in both qualitative and quantitative evaluations. Additionally, the enhanced results produced by our proposed IMIF method significantly improve the accuracy of object detection and keypoint detection. The is available at https://www.researchgate.net/publication/403951386_2026-IMIF.

Existing multi-color space guided techniques for underwater image enhancement (UIE) fail to take the advantages of the XYZ color space for preserving underwater image details, meanwhile, existing UIE datasets, typically containing low-quality reference images of distorted colors and blurred structures, lead to inaccurate enhancement mapping between low-quality and high-quality images. To overcome these above limitations, we propose a Multi-Color Space Fusion Network (MCSF-Net) for UIE. The MCSF-Net incorporates a Multi-dimensional Feature Fusion Block (MFFB) and weighted feature fusion scheme to effectively integrate complementary features from both XYZ and RGB color spaces. Moreover, we establish a Large-Scale Mixed UIE dataset (LSMU) by using nine no-reference metrics to filter out low-quality reference images from eight public UIE datasets, enabling more effective network learning. Extensive experiments on mainstream datasets demonstrate that the proposed method outperforms several leading approaches in both color restoration and detail enhancement of various underwater images. The code and dataset for MCSF-Net will be available athttps://github.com/WYJGR/MCSF-Net.

In hyperspectral image (HSI) classification, convolutional neural networks (CNNs) excel at local feature modeling but are limited to Euclidean space. Transformers offer long-range dependency modeling but suffer from high computational complexity. In contrast, graph convolutional networks (GCNs) can process information in non-Euclidean space, compensating for the limitations of CNNs. Meanwhile, the state space model Mamba, thanks to its linear complexity and strong long-range dependency modeling, shows great potential to offer an alternative to Transformers for HSI classification. To address the limitations of CNNs and Transformers while exploiting the potential of Mamba, we propose a dual-branch hybrid architecture named DBMGNet that combines Mamba with GCN for the HSI classification. In the Mamba branch, we design Band Selection Enhanced Bidirectional Mamba (BSEBM), which leverages Mamba’s long-range dependency modeling and sequential modeling capabilities to process spatial-spectral information. In the GCN branch, we apply reparameterized Chebyshev graph convolution to model similarity dependencies in non-Euclidean space, along with designing an adjacency matrix based on the intrinsic characteristics of HSIs. Extensive experiments demonstrate that our DBMGNet achieves the state-of-the-art performance of HSI classification against thirteen mainstream approaches. The code for this work will be available at https://github.com/Wanghao00pro/DBMGNet.

Time-of-flight magnetic resonance angiography (TOF-MRA) is a prevalent vascular imaging technique for assessing cerebrovascular diseases. Compared to routine 3T TOF-MRA, 7T TOF-MRA provides vascular structures with a higher signal-to-noise ratio (SNR) and better vessel contrast, revealing greater vascular details. However, the inaccessibility of 7T scanners and specific physiological and technical concerns limit its clinical application. Therefore, we aimed to generate high-quality 7T-like TOF-MRA from 3T TOF-MRA. Considering the spatial sparsity of vessel signals, the visibility discrepancy of distal and small vessels between 3T and 7T images, and the subtle spatial misalignment between paired data, we proposed a novel aleatoric-uncertainty-aware maximum intensity projection-based generative adversarial network (AU-MIPGAN). In our method, we employed a knowledge distillation (KD) framework to incorporate multi-directional MIP information into the 3T-to-7T learning process to strengthen the learning of vessels and provide three-dimensional (3D) vascular morphological knowledge for the student model, facilitating accurate generation of vascular structures. Furthermore, we exploited AU modeling to compensate for the spatial misalignment between paired 3T and 7T images during the training procedure, which helped the model concentrate more on learning the intrinsic gap between 3T and 7T images. Qualitative and quantitative results demonstrated that the proposed AU-MIPGAN can achieve promising performance for 7T-like TOF-MRA generation.

Underwater images widely endure severe color distortion because of the absorption and scattering of the water medium. We present a zero-approximation principle for underwater image color correction, called ZAP, to tackle this issue. Specifically, we first present the channel’s a and b pixel values to subtract the corresponding channels’ average pixel values so that the histograms corresponding to their pixel values are symmetric about the zero within the CIELab model. Afterward, we utilize the standard deviation of the channels’ a and b compensated pixel values to adjust the channels’ dynamic range and correct the image color distortion. Broad qualitative and quantitative experiments prove the practicability of ZAP for correcting color distortion in underwater images.

Underwater image quality is seriously degraded due to the insufficient light in water. Although artificial illumination can assist imaging, it often brings non-uniform illumination phenomenon. To this end, we develop an illumination channel sparsity prior (ICSP) guided variational framework for non-uniform illumination underwater image restoration. Technically, the illumination channel sparsity prior is built on the observation that the illumination channel of a uniform-light underwater image in HSI color space contains few pixels whose intensity is very low. Then according to the Retinex theory, we design a variational model with L0 norm term, constraint term, and gradient term, by integrating the proposed ICSP into an extended underwater image formation model. Such three regularizations are effective in enhancing the brightness, correcting color distortion, and revealing structures and fine-scale details. Meanwhile, we exploit a fast numerical algorithm on the base of the alternating direction method of multipliers (ADMM) to accelerate solving this optimization problem. We also collect a benchmark dataset, namely NUID that contains 925 real underwater images of different non-uniform illumination. Extensive experiments demonstrate that our proposed method is effective in terms of qualitative and quantitative comparisons, ablation studies, convergence analysis, and applications. The code and dataset are available athttps://github.com/Hou-Guojia/ICSP.

Due to the complex imaging mechanism, underwater images often suffer from multiple degradation issues, such as color cast, blurry detail, and low contrast, which affect the extraction of valuable information. To deal with these degradation issues, a simple yet effective underwater image quality improvement method based on color, detail and contrast restoration (CDCR) is developed, which consists of three key modules: a well-preserved finding-driven color balance module (CBM), a linear saturation transformation-based discriminant function-based detail restoration module (DRM), and a transmission minimization-oriented contrast restoration module (CRM). First, the CBM explores a well-preserved channel finding and employs a channel compensation strategy to balance the color differences among three color channels. Second, the DRM uses a piecewise underwater image saturation estimation strategy, which takes the various spectral properties of water into account and designs an additional linear saturation transformation-based discriminant function to prevent the transmission from being under-estimated. At last, the CRM estimates a global backscatter light based on transmission minimization and further improves the contrast by locally removing the backscatter light of the base layer. Our restored image is appealing in its natural color, fine details, and high contrast. Extensive experiments on three underwater image enhancement datasets show that our CDCR achieves better results than state-of-the-art methods, i.e., compared with the second-best method, the average PCQI and UIQM values of our method increase by 5.7% and 0.2%, and the average Blur and DFAD values of our method decrease by 8.0% and 5.3%. Meanwhile, experiments further suggest that the rate of new visible edges and the quality of contrast restoration of our CDCR at least increase by 7.7% and 51.2% in most tested sandstorm and foggy images, respectively, which demonstrates that our method has a good generalization capability for sandstorm and foggy image restoration.

Underwater imaging systems have evolved into essential hardware equipment for developing and utilizing marine resources. However, the complex underwater physical environment has often led to severe quality degradation of underwater visual perception. To address these issues, we design a principal component fusion method of foreground and background to enhance an underwater image, named PCFB. Specifically, we present a color balance-guided color correction strategy to remove color distortion issues that equalize the pixel values of the a and b channels of the CIELab color model. Subsequently, we implement a percentile maximum-based contrast enhancement strategy and a multilayer transmission map estimated dehazing strategy on the color-corrected image to yield the contrast-enhanced foreground and dehazed background sub-images. Finally, we employ a principal component analysis fusion method to reconstruct a high-visibility underwater image by integrating the advantages of the foreground contrast-enhanced sub-image and the background dehazed sub-image. Comprehensive experiments on three datasets demonstrate that our PCFB surpasses state-of-the-art methods both qualitatively and quantitatively. Moreover, our PCFB exhibits outstanding generalization capabilities for addressing haze and low-light images. The code is publicly available at:https://www.researchgate.net/publication/381259520_2024-PCFB.

Underwater images typically suffer from various quality degradation issues due to the scattering and absorption of light, but these degraded-quality underwater images are unbeneficial for analysis and applications. To effectively solve these quality degradation issues, an underwater image enhancement method via weighted wavelet visual perception fusion is introduced, called WWPF. Concretely, we first present an attenuation-map-guided color correction strategy to correct the color distortion of an underwater image. Subsequently, we employ the maximum information entropy optimized global contrast strategy to the color-corrected image to obtain a global contrast-enhanced image. Meanwhile, we apply a fast integration optimized local contrast strategy to the color-corrected image to get a local contrast-enhanced image. To exploit the complementary of the global contrast-enhanced image and the local contrast-enhanced image, we introduce a weighted wavelet visual perception fusion strategy to obtain a high-quality underwater image by fusing the high-frequency and low-frequency components of images at different scales. Our extensive experiments on three benchmarks validate that our WWPF outperforms the state-of-the-art methods in qualitative and quantitative. Besides, the underwater images processed by our WWPF also benefit practical underwater applications. The code is availablehttps://github.com/Li-Chongyi/WWPF_code.

The extraction of spatiotemporal neuron activity from calcium imaging videos plays a crucial role in unraveling the coding properties of neurons. While existing neuron extraction approaches have shown promising results, disturbing and scattering background and unused depth still impede their performance. To address these limitations, we develop an automatic and accurate neuron extraction paradigm, dubbed as decomposition–estimation–reconstruction (DER), consisting of D-procedure, E-procedure, and R-procedure. Specifically, the D-procedure first decomposes the raw data into a low-rank background and a sparse neuron signal, and regularizes$L_{0}$-norm priors of intensity and gradient of the neuron signal to suppress blurring and artifact effects. Then, the E-procedure estimates the depth-dependent transmission of the neuron signal based on its bright and dark channel priors. The R-procedure finally integrates the depth estimation of the neuron signal as a content-importance weight into a constrained non-negative matrix decomposition framework, which facilitates accurate neuron locations to boost the quality of extracted neurons. These three procedures are coupled in a cascade manner, where the former copes with calcium imaging data to facilitate the subsequent one. Comprehensive experiments on neuron extraction from calcium imaging videos demonstrate the superiority of our DER paradigm in both qualitative results and quantitative assessments over state-of-the-art methods.

Remote sensing images are significantly degraded by bad weather conditions, such as haze and sandstorms, which provide unhelpful support for valuable information extraction. Most existing remote sensing image enhancement methods ignore the wavelength dependence of the scattering coefficient and local scattering differences of images, and therefore cannot well handle the colorized haze in which the medium transmission varies in different color channels. In this article, we propose a single image quality enhancement method using joint local structure dehazing and local texture enhancement (SDTE). Specifically, SDTE first uses a minimal channel between r, g, and b channels to estimate a coarse local airlight, and designs an achromatic airlight-driven refinement strategy to refine it. Meanwhile, SDTE estimates a local transmission via independent calculation of r, g, and b channels, which tackles the limitation that existing methods heavily depend on global transmission over the entire image. Then, SDTE removes the haze and amplifies the gradient using the estimated local airlight and transmission, thereby preserving significant structures and enhancing fine details. Finally, SDTE introduces an adaptive color correction based on the ranking of channel mean value and two channel-dependent gain factors to further eliminate the severe color distortion. More specially, we also collect a remote sensing colorized hazy image enhancement benchmark (RSCHI) including 339 remote sensing images captured in colorized haze or sandstorm, which makes it pay more attention to the color cast issue. We conduct a comprehensive study on benchmark datasets of RSCHI and UIEB and indicate better performance than the state-of-the-art (SOTA) methods. Meanwhile, we use a series of ablation studies to demonstrate the effectiveness and robustness of each key contribution and validate its generalization performance in other scenes.

Wheat variety identification from hyperspectral images holds significant importance in both fine breeding and intelligent agriculture. However, the discriminatory accuracy of some techniques is limited due to insufficient datasets, data redundancy, and noise interference. To address these issues, we propose a wheat variety identification framework called generate adversarial-driven cross-aware network (GACNet), comprising a semi-supervised generative adversarial network (GAN) for data augmentation and a cross-aware attention network (CAANet) for variety identification. First, the semi-supervised GAN (SSGAN) alleviates data scarcity by generating fake hyperspectral images as realistically as possible through learning the distribution hypothesis of real hyperspectral images, while the discriminator distinguishes between real and fake hyperspectral images. Subsequently, the CAANet is employed for wheat variety identification, which leverages a cascading cross-learning of 3-D and 2-D convolutions to fully exploit spectral, spatial, and texture features and refines the features through an embedded attention mechanism in the cross-convolutional module. Additionally, we constructed a hyperspectral wheat variety dataset (HWVD) comprising 4560 samples of 19 categories. Extensive experiments on our dataset demonstrate that our GACNet outperforms state-of-the-art methods for wheat variety identification. The HWVD will be made available.

This letter proposes an effective underwater image restoration method that consists of a local image descattering and an adaptive brightness improvement. First, we establish an adaptive objective function for improving the brightness of underwater image according to the best-preserved channel of an image, and an augmented Lagrange multiplier based alternating direction minimization algorithm is derived to solve the optimization problem. Second, we introduce a local transmission estimation method that takes into account the different attenuation of light on the red, green and blue channels, which overcomes the limitation that existing methods heavily depend on the global transmission over the entire image. Extensive experiments on real-world underwater images demonstrate the effectiveness of the proposed method in underwater image restoration. Moreover, our method shows good generalization capability for enhancing remote sensing and nighttime images.

Due to the absorption and scattering of light, underwater captured images often face serious quality degradation issues. In this letter, we propose to cope with the aforementioned issues via piecewise color correction and dual prior optimized contrast enhancement. Specifically, we first present the piecewise color correction method using the maximum mean and two gain factors to correct the color cast of each color channel. Then, we propose a dual prior optimized contrast enhancement method, which relies on the spatial and texture priors to decompose the base layer and detail layer of the V channel in HSV color space. Meanwhile, we employ different enhancement strategies in different layers to enhance the contrast and texture detail of underwater images. Our extensive experiments on several benchmark datasets show that our method outperforms eleven compared state-of-the-art methods. Moreover, our method has good generalization capability for fog and low-light images. The code is available athttps://github.com/Li-Chongyi/PCDE.

Self-supervised representation learning (SSL) typically suffers from inadequate data utilization and feature-specificity due to the suboptimal sampling strategy and the monotonous optimization method. Existing contrastive-based methods alleviate these issues through exceedingly long training time and large batch size, resulting in non-negligible computational consumption and memory usage. In this paper, we present an efficient self-supervised framework, called GLNet. The key insights of this work are the novel sampling and ensemble learning strategies embedded in the self-supervised framework. We first propose a location-based sampling strategy to integrate the complementary advantages of semantic and spatial characteristics. Whereafter, a Siamese network with momentum update is introduced to generate representative vectors, which are used to optimize the feature extractor. Finally, we particularly embed global contrastive and local location tasks in the framework, which aims to leverage the complementarity between the high-level semantic features and low-level texture features. Such complementarity is significant for mitigating the feature-specificity and improving the generalizability, thus effectively improving the performance of downstream tasks. Extensive experiments on representative benchmark datasets demonstrate that GLNet performs favorably against the state-of-the-art SSL methods. Specifically, GLNet improves MoCo-v3 by 2.4% accuracy on ImageNet dataset, while improves 2% accuracy and consumes only 75% training time on the ImageNet-100 dataset. In addition, GLNet is appealing in its compatibility with popular SSL frameworks. Code is available at GLNet.

We develop a novel depth-estimation constrained nonnegative matrix factorization framework for reinforcing neuron extraction, which takes advantage of accurate depth estimation to alleviate the scattering of raw data and compensate for depth loss in the vectorization operation. Established on bright and dark channel priors, a depth-dependent transmission estimation model is effective in estimating the more accurate depth of raw data, where the atmospheric light of each pixel is finely estimated to alleviate the problem of image descattering and overcome the limitation of conventionally constant assumption. Besides, our framework is simply implemented in constrained nonnegative matrix factorization, and can be flexibly accommodated to various neuron extraction approaches. Extensive experiments confirm the superior performance of our framework in terms of reinforcing more accuracy of neuron extraction and demixing better results of spatially overlapping neurons, moreover, the utility of our depth-estimation model is proved for imaging whole brain of zebrafish larvae.

Currently, most existing methods using hyperspectral image to assist seed identification only consider the spectral information but ignore the spatial information resulting in unsatisfactory classification results. To cope with this issue, we propose a spatial, spectral, and texture-aware attention network to identify corn varieties, called SSTNet. Specifically, we first employ 3D convolution to extract the spatial and inter-spectral features. Subsequently, we utilize 2D convolution to extract the spatial and texture features. Meanwhile, we embed an attention mechanism into the 2D convolution module to further refine the spatial and texture features. The advantageous complementary properties of 3D and 2D convolutions allow the spatial and textural features of hyperspectral images to be fully exploited. Besides, we construct a hyperspectral image dataset including 1200 samples of 10 corn varieties. Experiments on our proposed dataset demonstrate that our SSTNet outperforms the state-of-the-art methods for identifying corn varieties.

Recently, image restoration has attracted considerable attention from researchers, and these methods generally restore degraded images based on the atmospheric scattering model (ATSM) and retinex model (RM). The two models only take into the single attenuation process during imaging, thereby introducing undesirable results. To deal with this issue, we propose an image restoration method based on a generalized image formation model (GIFM). First, unlike the existing image restoration methods, we rebuild a novel image formation model, which describes the light attenuation process that includes the light source-scene path and scene-sensor path. Second, we construct an objective optimization function to decompose a degraded image into a color distorted component and color corrected component, and an augmented Lagrange multiplier-based alternating direction minimization algorithm is provided to solve the optimization problem. Finally, we fully consider the advantages of the small-scale neighborhood and large-scale neighborhood in image restoration, and an image itself brightness-based weighted fusion strategy is proposed to balance brightness enhancement and contrast improvement. Extensive experiments on three image enhancement datasets show that our GIFM achieves better results than state-of-the-art methods. Experiments further suggest that our GIFM performs well for image restoration of extreme scenes, keypoint detection, object detection, and image segmentation.

Underwater images typically suffer from color deviations and low visibility due to the wavelength-dependent light absorption and scattering. To deal with these degradation issues, we propose an efficient and robust underwater image enhancement method, called MLLE. Specifically, we first locally adjust the color and details of an input image according to a minimum color loss principle and a maximum attenuation map-guided fusion strategy. Afterward, we employ the integral and squared integral maps to compute the mean and variance of local image blocks, which are used to adaptively adjust the contrast of the input image. Meanwhile, a color balance strategy is introduced to balance the color differences between channel a and channel b in the CIELAB color space. Our enhanced results are characterized by vivid color, improved contrast, and enhanced details. Extensive experiments on three underwater image enhancement datasets demonstrate that our method outperforms the state-of-the-art methods. Our method is also appealing in its fast processing speed within 1s for processing an image of size 1024×1024×3 on a single CPU. Experiments further suggest that our method can effectively improve the performance of underwater image segmentation, keypoint detection, and saliency detection. The project page is available at https://li-chongyi.github.io/proj_MMLE.html.

Underwater image enhancement aims at improving the visibility and eliminating color distortions of underwater images degraded by light absorption and scattering in water. Recently, retinex variational models show remarkable capacity of enhancing images by estimating reflectance and illumination in a retinex decomposition course. However, ambiguous details and unnatural color still challenge the performance of retinex variational models on underwater image enhancement. To overcome these limitations, we propose a hyper-laplacian reflectance priors inspired retinex variational model to enhance underwater images. Specifically, the hyper-laplacian reflectance priors are established with thel1/2-norm penalty on first-order and second-order gradients of the reflectance. Such priors exploit sparsity-promoting and complete-comprehensive reflectance that is used to enhance both salient structures and fine-scale details and recover the naturalness of authentic colors. Besides, thel2norm is found to be suitable for accurately estimating the illumination. As a result, we turn a complex underwater image enhancement issue into simple subproblems that separately and simultaneously estimate the reflection and the illumination that are harnessed to enhance underwater images in a retinex variational model. We mathematically analyze and solve the optimal solution of each subproblem. In the optimization course, we develop an alternating minimization algorithm that is efficient on element-wise operations and independent of additional prior knowledge of underwater conditions. Extensive experiments demonstrate the superiority of the proposed method in both subjective results and objective assessments over existing methods.

We develop a variational retinex algorithm for enhancing single underwater image with multiorder gradient priors of reflectance and illumination. First, a simple yet effective color correction approach is used to remove color casts and recover naturalness. Then, a variational retinex model for enhancing the color-corrected underwater image is established by imposing multiorder gradient priors of reflectance and illumination. According to structural sparsity difference between illumination and reflectance, the l1norm is accurately adopted to model piecewise and piecewise linear approximations on the reflectance, while the l2norm is appropriately employed to enforce spatial smoothness and spatial linear smoothness on the illumination. Next, a complex underwater image enhancement issue is turned into simple denoising subproblems, which can be addressed by an efficient optimization algorithm that is fast performed on pixel-wise operations without requiring additional prior knowledge about underwater imaging conditions. Final experiments demonstrate that the proposed method yields better results of qualitative and quantitative assessments than several traditional and leading underwater image enhancement approaches.

We develop a simple yet effective model for enhancing single underwater image by mathematically extending the Beer-Lambert law. In the proposed model, we take advantage of the mean and variance of natural images to be the reference to correct color casts of underwater images. We propose an efficient strategy to recover better details of underwater images, which involves two steps: in the first step we establish a linear model associated with the mean and variance of underwater images to locate images regions containing more details, and in the second step we present a nonlinear adaptive weight scheme using this locating information to recover better details and prevent partial over-enhancement. Ultimate experiments are performed to demonstrate the effectiveness of the proposed method, and these experimental results show that our method yields better structural restoration, more naturalness color correction, and less time consumption.s.

The extreme channels prior (ECP) relies on the bright and dark channels of an image, and the corresponding ECP-based methods perform well in blind image deblurring. However, we experimentally observe that the pixel values of dark and bright channels in some images are not concentratedly distributed on 0 and 1 respectively. Based on this observation, we develop a model with a joint prior which combines the extreme channels prior and the L0-regularized intensity and gradient prior for blind image deblurring, and previous image deblurring approaches based on dark channel prior, L0- regularized intensity and gradient, and extreme channels prior can be seen as a particular case of our model. Then we derive an efficient optimization algorithm using the half-quadratic splitting method to address the non-convex L0-minimization problem. A large number of experiments are finally performed to demonstrate the superiority of the proposed model in details restoration and artifacts removal, and our model outperforms several leading deblurring approaches in terms of subjective results and objective assessments. In addition, our method is more applicable for deblurring natural, text and face images which do not contain many bright or dark pixels.

We develop a novel compressed sensing magnetic resonance imaging (CSMRI) algorithm with joint image and patch priors, where the total variation (TV) is adopted for image-level sparse prior and the expected patch log likelihood (EPLL) is used for patch-level sparse prior. The proposed joint priors capture global and local sparse nature of the MR image to promote image structures and suppress artifacts or noise, which can alleviate the limitations of previous CSMRI methods. And we derive an appropriate cost function which can be addressed by an efficient optimization scheme that iteratively alternates among l1norm approximation, latent patch reconstruction and ideal image reconstruction. Final experiments are provided to show the satisfactory performance of the proposed method in MRI reconstruction, which outperforms other competitive CSMRI approaches in both subjective results and objective assessments.

Since the light is absorbed and scattered while traveling in water, color distortion, under-exposure and fuzz are three major problems of underwater imaging. In this paper, a novel retinex-based enhancing approach is proposed to enhance single underwater image. The proposed approach has mainly three steps to solve the problems mentioned above. First, a simple but effective color correction strategy is adopted to address the color distortion. Second, a variational framework for retinex is proposed to decompose the reflectance and the illumination, which represent the detail and brightness respectively, from single underwater image. An effective alternating direction optimization strategy is adopted to solve the proposed model. Third, the reflectance and the illumination are enhanced by different strategies to address the under-exposure and fuzz problem. The final enhanced image is obtained by combining use the enhanced reflectance and illumination. The enhanced result is improved by color correction, lightens dark regions, naturalness preservation, and well enhanced edges and details. Moreover, the proposed approach is a general method that can enhance other kinds of degraded image, such as sandstorm image.

This paper proposes a novel non-parametric Bayesian framework for solving mixed noise removal problem. In order to removing unstable effects of outlier noise such as salt-and-pepper in the training data, we decompose the observed data model into three components terms of ideal data, Gaussian noise and sparse outlier. And the proposed model employs spike-slab sparse prior to find the sparser coefficients of desired data term and outlier noise. Note that the proposed non-parametric Bayesian model can infer the noise statistics from the training data and have been robust to the mixed noise without tuning of model parameters. Experimental results demonstrate our proposed algorithm performs well with mixed noise and achieves better performance over other state-of-the-art methods.