Hazy Images Research Articles

WTCL-Dehaze: Rethinking Real-World Image Dehazing via Wavelet Transform and Contrastive Learning

Images captured in hazy outdoor conditions often suffer from colour distortion, low contrast, and loss of detail, which impair high-level vision tasks. Single image dehazing is essential for applications such as autonomous driving and surveillance, with the aim of restoring image clarity. In this work, we propose WTCL-Dehaze an enhanced semi-supervised dehazing network that integrates Contrastive Loss and Discrete Wavelet Transform (DWT). We incorporate contrastive regularization to enhance feature representation by contrasting hazy and clear image pairs. Additionally, we utilize DWT for multi-scale feature extraction, effectively capturing high-frequency details and global structures. Our approach leverages both labelled and unlabelled data to mitigate the domain gap and improve generalization. The model is trained on a combination of synthetic and real-world datasets, ensuring robust performance across different scenarios. Extensive experiments demonstrate that our proposed algorithm achieves superior performance and improved robustness compared to state-of-the-art single image dehazing methods on both benchmark datasets and real-world images.

Image dehazing via self-supervised depth guidance

Self-supervised learning methods have demonstrated promising benefits to feature representation learning for image dehazing tasks, especially for avoiding the laborious work of collecting hazy-clean image pairs, while also enabling better generalization abilities of the model. Despite the long-standing interests in depth estimation for image dehazing tasks, few works have fully explored the interactions between depth and dehazing tasks in an unsupervised manner. In this paper, a self-supervised image dehazing framework under the guidance of self-supervised depth estimation has been proposed, to fully exploit the interactions between depth and hazes for image dehazing. Specifically, the hazy image and the corresponding depth estimation are generated and optimized from the clear image in a dual-network self-supervised manner. The correlations between depth and hazy images are exploited in depth-guided hybrid attention Transformer blocks, which adaptively leverage both the cross-attention and self-attention to effectively model hazy densities via cross-modality fusion and capture global context information for better feature representations. In addition, the depth estimations of hazy images are further explored for the detection tasks on hazy images. Extensive experiments demonstrate that the depth estimation not only enhances the model generalization ability across different dehazing datasets, leading to state-of-the-art self-supervised dehazing performance, but also benefits downstream detection tasks on hazy images. Our code is available at https://github.com/DongLiangSXU/Depth-Guidance-dehazing.git.

Degradation Type-Aware Image Restoration for Effective Object Detection in Adverse Weather

Despite significant advancements in CNN-based object detection technology, adverse weather conditions can disrupt imaging sensors’ ability to capture clear images, thereby adversely impacting detection accuracy. Mainstream algorithms for adverse weather object detection enhance detection performance through image restoration methods. Nevertheless, the majority of these approaches are designed for a specific degradation scenario, making it difficult to adapt to diverse weather conditions. To cope with this issue, we put forward a degradation type-aware restoration-assisted object detection network, dubbed DTRDNet. It contains an object detection network with a shared feature encoder (SFE) and object detection decoder, a degradation discrimination image restoration decoder (DDIR), and a degradation category predictor (DCP). In the training phase, we jointly optimize the whole framework on a mixed weather dataset, including degraded images and clean images. Specifically, the degradation type information is incorporated in our DDIR to avoid the interaction between clean images and the restoration module. Furthermore, the DCP makes the SFE possess degradation category awareness ability, enhancing the detector’s adaptability to diverse weather conditions and enabling it to furnish requisite environmental information as required. Both the DCP and the DDIR can be removed according to requirement in the inference stage to retain the real-time performance of the detection algorithm. Extensive experiments on clear, hazy, rainy, and snowy images demonstrate that our DTRDNet outperforms advanced object detection algorithms, achieving an average mAP of 79.38% across the four weather test sets.

Multiscale implicit frequency selective network for single‐image dehazing

AbstractImage dehazing is aimed to reconstruct a clear latent image from a degraded image affected by haze. Although vision transformers have achieved impressive success in various computer vision tasks, the limitations in scale and quality of available datasets have hindered the transformer effectiveness for image dehazing. Thus, convolutional neural networks (CNNs) remain the mainstream approach for image dehazing, offering robust performance and adaptability. We further explore the potential of CNNs in image dehazing by proposing a multiscale implicit frequency selection network (MIFSN). The proposed MIFSN enhances multiscale representation learning based on U‐shaped networks. As hazy and clear images considerably differ in high‐frequency components, we introduce an implicit frequency selection module to amplify high‐frequency components of features and generate candidate feature maps. Implicit frequency selection attention is then used to emphasize and merge beneficial frequency components. Results from extensive experiments on synthetic and real‐world datasets demonstrate the superior performance of MIFSN for image dehazing.

Adaptive haze pixel intensity perception transformer structure for image dehazing networks.

In the realm of deep learning-based networks for dehazing using paired clean-hazy image datasets to address complex real-world haze scenarios in daytime environments and cross-dataset challenges remains a significant concern due to algorithmic inefficiencies and color distortion. To tackle these issues, we propose SwinTieredHazymers (STH), a dehazing network designed to adaptively discern pixel intensities in hazy images and compute haze residue for clarity restoration. Through a unique three-branch design, we hierarchically modulate haze residuals by leveraging the global features brought by Transformer and the local features brought by Convolutional Neural Network (CNN) which has led to the algorithm's widespread applicability. Experimental results demonstrate that our approach surpasses advanced single-image dehazing methods in both quantitative metrics and visual fidelity for real-world hazy image dehazing, while also exhibiting strong performance in cross-dataset dehazing scenarios.

Retina-Inspired Lightweight Spiking Convolutional Neural Network for Single-Image Dehazing.

Suspended particles in hazy medium absorb and scatter light, severely degrading imaging quality. Numerous single-image dehazing methods have been proposed to reconstruct clear images from hazy ones. However, most of them focus on increasing depth and width to improve dehazing performance, which incurs high computation and energy costs. To address this issue, we propose a lightweight spiking convolutional neural network (CNN) referred to as retina-inspired spiking CNN (RI-SCNN) for the reconstruction of hazy images. Unlike conventional dehazing techniques, first, our proposed network simulates the hierarchical structure and cellular function of the retina and devises five network modules to efficiently encode and extract image features through ON and OFF roads. Furthermore, the linear reconstruction mechanism is introduced to integrate the outputs from different roads, adaptively preserving regions with optimal details and constructing a comprehensive visual representation. Finally, by the transformed atmospheric scattering formula, our network can generate the dehazy image. Incorporating the microscale spiking mechanism of the brain, the entire network leverages discrete binary spike trains for information encoding and transmission, directly trained by spiking surrogate gradient learning on integrate-and-fire (IF) neurons. Experimental results demonstrate the superiority of the proposed RI-SCNN in terms of quantitative dehazing performance, qualitative visual effect, energy efficiency, and run speed. Considering its lightweight architecture with ultralow computation and energy costs, the network is encouraged to be deployed in the visual sensor hardware to improve overall performance.

A novel multiscale cGAN approach for enhanced salient object detection in single haze images

In computer vision, image dehazing is a low-level task that employs algorithms to analyze and remove haze from images, resulting in haze-free visuals. The aim of Salient Object Detection (SOD) is to locate the most visually prominent areas in images. However, most SOD techniques applied to visible images struggle in complex scenarios characterized by similarities between the foreground and background, cluttered backgrounds, adverse weather conditions, and low lighting. Identifying objects in hazy images is challenging due to the degradation of visibility caused by atmospheric conditions, leading to diminished visibility and reduced contrast. This paper introduces an innovative approach called Dehaze-SOD, a unique integrated model that addresses two vital tasks: dehazing and salient object detection. The key novelty of Dehaze-SOD lies in its dual functionality, seamlessly integrating dehazing and salient object identification into a unified framework. This is achieved using a conditional Generative Adversarial Network (cGAN) comprising two distinct subnetworks: one for image dehazing and another for salient object detection. The first module, designed with residual blocks, Dark Channel Prior (DCP), total variation, and the multiscale Retinex algorithm, processes the input hazy images. The second module employs an enhanced EfficientNet architecture with added attention mechanisms and pixel-wise refinement to further improve the dehazing process. The outputs from these subnetworks are combined to produce dehazed images, which are then fed into our proposed encoder–decoder framework for salient object detection. The cGAN is trained with two modules working together: the generator aims to produce haze-free images, whereas the discriminator distinguishes between the generated haze-free images and real haze-free images. Dehaze-SOD demonstrates superior performance compared to state-of-the-art dehazing methods in terms of color fidelity, visibility enhancement, and haze removal. The proposed method effectively produces high-quality, haze-free images from various hazy inputs and accurately detects salient objects within them. This makes Dehaze-SOD a promising tool for improving salient object detection in challenging hazy conditions. The effectiveness of our approach has been validated using benchmark evaluation metrics such as mean absolute error (MAE), peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM).

PSD-ELGAN: A pseudo self-distillation based CycleGAN with enhanced local adversarial interaction for single image dehazing

Compared to pixel-level content loss, domain-level style loss in CycleGAN-based dehazing algorithms just imposes relatively soft constraints on the intermediate translated images, resulting in struggling to accurately model haze-free features from real hazy scenes. Furthermore, globally perceptual discriminator may misclassify real hazy images with significant scene depth variations as clean style, thereby resulting in severe haze residue. To address these issues, we propose a pseudo self-distillation based CycleGAN with enhanced local adversarial interaction for image dehazing, termed as PSD-ELGAN. On the one hand, we leverage the characteristic of CycleGAN to generate pseudo image pairs during training. Knowledge distillation is employed in this unsupervised framework to transfer the informative high-quality features from the self-reconstruction network of real clean images to the dehazing generator of paired pseudo hazy images, which effectively improves its haze-free feature representation ability without increasing network parameters. On the other hand, in the output of dehazing generator, four non-uniform image patches severely affected by residual haze are adaptively selected as input samples. The local discriminator could easily distinguish their hazy style, thereby further compelling the dehazing generator to suppress haze residues in such regions, thus enhancing its dehazing performance. Extensive experiments show that our PSD-ELGAN can achieve promising results and better generality across various datasets.

Fast no-reference deep image dehazing

This paper presents a deep learning method for image dehazing and clarification. The main advantages of the method are high computational speed and using unpaired image data for training. The method adapts the Zero-DCE approach (Li et al. in IEEE Trans Pattern Anal Mach Intell 44(8):4225–4238, 2021) for the image dehazing problem and uses high-order curves to adjust the dynamic range of images and achieve dehazing. Training the proposed dehazing neural network does not require paired hazy and clear datasets but instead utilizes a set of loss functions, assessing the quality of dehazed images to drive the training process. Experiments on a large number of real-world hazy images demonstrate that our proposed network effectively removes haze while preserving details and enhancing brightness. Furthermore, on an affordable GPU-equipped laptop, the processing speed can reach 1000 FPS for images with 2K resolution, making it highly suitable for real-time dehazing applications.

Multi-scale large kernel convolution and hybrid attention network for remote sensing image dehazing

Remote sensing (RS) image dehazing holds significant importance in enhancing the quality and information extraction capability of RS imagery. The enhancement in image dehazing quality has progressively advanced alongside the evolution of convolutional neural network (CNN). Due to the fixed receptive field of CNN, there is insufficient utilization of contextual information on haze features in multi-scale RS images. Additionally, the network fails to adequately extract both local and global information of haze features. In addressing the above problems, in this paper, we propose an RS image dehazing network based on multi-scale large kernel convolution and hybrid attention (MKHANet). The network is mainly composed of multi-scale large kernel convolution (MSLKC) module, hybrid attention (HA) module and feature fusion attention (FFA) module. The MSLKC module fully fuses the multi-scale information of features while enhancing the effective receptive field of the network by parallel multiple large kernel convolutions. To alleviate the problem of uneven distribution of haze and effectively extract the global and local information of haze features, the HA module is introduced by focusing on the importance of haze pixels at the channel level. The FFA module aims to boost the interaction of feature information between the network's deep and shallow layers. The subjective and objective experimental results on on multiple RS hazy image datasets illustrates that MKHANet surpasses existing state-of-the-art (SOTA) approaches. The source code is available at https://github.com/tohang98/MKHA_Net.

A Lightweight CNN Based on Axial Depthwise Convolution and Hybrid Attention for Remote Sensing Image Dehazing

Hazy weather reduces contrast, narrows the dynamic range, and blurs the details of the remote sensing image. Additionally, color fidelity deteriorates, causing color shifts and image distortion, thereby impairing the utility of remote sensing data. In this paper, we propose a lightweight remote sensing-image-dehazing network, named LRSDN. The network comprises two tailored, lightweight modules arranged in cascade. The first module, the axial depthwise convolution and residual learning block (ADRB), is for feature extraction, efficiently expanding the convolutional receptive field with little computational overhead. The second is a feature-calibration module based on the hybrid attention block (HAB), which integrates a simplified, yet effective channel attention module and a pixel attention module embedded with an observational prior. This joint attention mechanism effectively enhances the representation of haze features. Furthermore, we introduce a novel method for remote sensing hazy image synthesis using Perlin noise, facilitating the creation of a large-scale, fine-grained remote sensing haze image dataset (RSHD). Finally, we conduct both quantitative and qualitative comparison experiments on multiple publicly available datasets. The results demonstrate that the LRSDN algorithm achieves superior dehazing performance with fewer than 0.1M parameters. We also validate the positive effects of the LRSDN in road extraction and land cover classification applications.

Enhancing low-light images via dehazing principles: Essence and method

Given the visual resemblance between inverted low-light and hazy images, dehazing principles are borrowed to enhance low-light images. However, the essence of such methods remains unclear, and they are susceptible to over-enhancement. Regarding the above issues, in this letter, we present corresponding solutions. Specifically, we point out that the Haze Formation Model (HFM) used for image dehazing exhibits a Bidirectional Mapping Property (BMP), enabling adjustment of image brightness and contrast. Building upon this property, we give a comprehensive and in-depth theoretical explanation for why dehazing on inverted low-light image is a solution to the image brightness enhancement problem. Further, an Adaptive Full Dynamic Range Mapping (AFDRM) method is then proposed to guide HFM in restoring the visibility of low-light images without inversion, while overcoming the issue of over-enhancement. Extensive experiments validate our proof and demonstrate the efficacy of our method.

Robust polarimetric dehazing algorithm based on low-rank approximation and multiple virtual-exposure fusion

Polarimetric dehazing is an effective way to enhance the quality of images captured in foggy weather. However, images of essential polarization parameters are vulnerable to noise, and the brightness of dehazed images is usually unstable due to different environmental illuminations. These two weaknesses reveal that current polarimetric dehazing algorithms are not robust enough to deal with different scenarios. This paper proposes a novel, to our knowledge, and robust polarimetric dehazing algorithm to enhance the quality of hazy images, where a low-rank approximation method is used to obtain low-noise polarization parameter images. Besides, in order to improve the brightness stability of the dehazed image and thus keep the image have more details within the standard dynamic range, this study proposes a multiple virtual-exposure fusion (MVEF) scheme to process the dehazed image (usually having a high dynamic range) obtained through polarimetric dehazing. Comparative experiments show that the proposed dehazing algorithm is robust and effective, which can significantly improve overall quality of hazy images captured under different environments.

Unpaired Remote Sensing Image Dehazing Using Enhanced Skip Attention-Based Generative Adversarial Networks with Rotation Invariance

Remote sensing image dehazing aims to enhance the visibility of hazy images and improve the quality of remote sensing imagery, which is essential for various applications such as object detection and classification. However, the lack of paired data in remote sensing image dehazing enhances the applications of unpaired image-to-image translation methods. Nonetheless, the considerable parameter size of such methods often leads to prolonged training times and substantial resource consumption. In this work, we propose SPRGAN, a novel approach leveraging Enhanced Perlin Noise-Based Generative Adversarial Networks (GANs) with Rotation Invariance to address these challenges. Firstly, we introduce a Spatial-Spectrum Attention (SSA) mechanism with Skip-Attention (SKIPAT) to enhance the model’s ability to interpret and process spectral information in hazy images. Additionally, we have significantly reduced computational overhead to streamline processing. Secondly, our approach combines Perlin Noise Masks in pre-training to simulate real foggy conditions, thereby accelerating convergence and enhancing performance. Then, we introduce a Rotation Loss (RT Loss) to ensure the model’s ability to dehaze images from different angles uniformly, thus enhancing its robustness and adaptability to diverse scenarios. At last, experimental results demonstrate the effectiveness of SPRGAN in remote sensing image dehazing, achieving better performance compared to state-of-the-art methods.

Learning depth-aware decomposition for single image dehazing

Image dehazing under deficient data is an ill-posed and challenging problem. Most existing methods tackle this task by developing either CycleGAN-based hazy-to-clean translation or physical-based haze decomposition. However, geometric structure is often not effectively incorporated in their straightforward hazy-clean projection framework, which might incur inaccurate estimation in distant areas. In this paper, we rethink the image dehazing task and propose a depth-aware perception framework, DehazeDP, for robust haze decomposition on deficient data. Our DehazeDP is insthe pired by Diffusion Probabilistic Model to form an end-to-end training pipeline that seamlessly ines the hazy image generation with haze disentanglement. Specifically, in the forward phase, the haze is added to a clean image step-by-step according to the depth distribution. Then, in the reverse phase, a unified U-Net is used to predict the haze and recover the clean image progressively. Extensive experiments on public datasets demonstrate that the proposed DehazeDP performs favorably against state-of-the-art approaches. We release the code and models at https://github.com/stallak/DehazeDP.

Unsupervised multi-branch network with high-frequency enhancement for image dehazing

Recently, CycleGAN-based methods have been widely applied to the unsupervised image dehazing and achieved significant results. However, most existing CycleGAN-based methods ignore that the input of the generator contains two different distributions of data which can often lead to confusion in the learning process of the generator, consequently limiting the final dehazing performance. Moreover, reconstructing clear images through model architecture design and loss functions is an indirect constraint, making it difficult to compensate for the missing high-frequency information, such as textures and structures in the extracted features from hazy images. To address these issues, in this paper, we propose an Unsupervised Multi-Branch with High-Frequency Enhancement Network (UME-Net) which contain an Multi-Branch Dehazing Network (MBDN) and a High-Frequency Components Enhancement Module (HFEM). Specifically, MBDN constructs a single unsupervised dehazing network with Shared Encoding Module (SEM) and Multi-Branch Decoding Module (MDM). SEM enhance the consistency of feature representation and MDM effectively addresses the confusion during the generator learning process in CycleGAN-based methods. Furthermore, based on a key observation that hazy images and their corresponding clear images exhibit only subtle differences in high-frequency information, the HFEM is designed to compensates for the missing high-frequency information in the network which further enhances the reconstruction capability of the UME-Net for restore edge and texture information in obscured by dense haze. Experimental results on challenging benchmark datasets demonstrate the superiority of our UME-Net over SOTA unsupervised image dehazing methods. The source code is available at https://www.github.com/thislzm/UME-Net.

Ship polarization information restoration in sea haze weather via airlight estimation

ABSTRACT The polarization information of incident light gets lost after passing through the scattering medium such as haze. This phenomenon, called depolarization, poses a challenge to polarization remote sensing of ships in sea-haze weather. In this paper, we propose the concept of ship polarization information restoration for the first time and propose a ship polarization information restoration method for the benefit of ship remote sensing in sea haze weather. The approach is based on the estimation of airlight, a component of hazy polarimetric images. To demonstrate the effectiveness of the proposed ship polarization information restoration method, we conduct an outdoor experiment on ship polarization information restoration in sea haze weather. Experimental results show that sea haze significantly degrades the degree of polarization of ships, but slightly affects the angle of polarization. In addition, the degree of polarization of ships becomes more prominent after polarization information restoration.

Pythagorean Fuzzy Set for Enhancement of Low Contrast Mammogram Images

ABSTRACTBreast masses are often one of the primary signs of breast cancer, and precise segmentation of these masses is essential for accurate diagnosis and treatment planning. Diagnosis may be complex depending on the size and visibility of the mass. When the mass is not visible clearly, precise segmentation becomes very difficult and in that case enhancement is essential. Inadequate compression, patient movement, or paddle/breast movement during the exposure process might cause hazy mammogram images. Without enhancement, accurate segmentation and detection cannot be done. As there exists uncertainties in different regions, reducing uncertainty is still a main problem and so fuzzy methods may deal these uncertainties in a better way. Though there are many fuzzy and advanced fuzzy methods, we consider Pythagorean fuzzy set as one of the fuzzy sets that may be powerful to deal with uncertainty. This research proposes a new Pythagorean fuzzy methodology for mammography image enhancement. The image is first transformed into a fuzzy image, and the nonmembership function is then calculated using a newly created Pythagorean fuzzy generator. Membership function of Pythagorean fuzzy image is computed from nonmembership function. The plot between the membership value and the hesitation degree is used to calculate a constant term in the membership function. Next, an enhanced image is obtained by applying fuzzy intensification operator to the Pythagorean fuzzy image. The proposed method is compared qualitatively and quantitatively with those of non‐fuzzy, intuitionistic fuzzy, Type 2 fuzzy, and Pythagorean fuzzy methods, it is found that the suggested method outperforms the other methods. To show the usefulness of the proposed enhanced method, segmentation is carried out on the enhanced images.

Classification and removal of hazy images based on a transmission fusion strategy using the Alexnet network

Classification and removal of hazy images based on a transmission fusion strategy using the Alexnet network

Single Image Dehazing via Regional Saturation-Value Translation

This paper proposes an image dehazing prior, called Regional Saturation-Value Translation (RSVT), in order to address the color distortion issues produced by prevailing prior-based dehazing methods when processing hazy images with large sky regions. The proposed RSVT prior is derived from statistical analyses of the correlation between hazy points and respective haze-free points in the HSV color space. The prior is based upon two key observations in the sky areas. First, the difference in terms of hue for a pair of hazy and haze-free points is very small, raising an assumption that the variability of pixel values caused by haze mostly occurs in the saturation and value spaces. This leads to the second observation that, in the 2D saturation-value coordinate system, almost all the lines passing through corresponding pairs of hazy-clean points, termed S-V lines, are likely to intersect around the airlight coordinates. A hybrid refined dark channel is introduced in order to decompose the input hazy image into sky and non-sky regions and to estimate the global atmospheric light. Combining the prior with the hybrid refined dark channel, a novel single image dehazing framework is proposed. Haze removal is performed separately for the sky and non-sky regions by adopting the proposed RSVT prior and Koschmieder's law, respectively. The experimental results have shown that the proposed dehazing method can restore visually compelling sky color and effectively handle the color distortion issues associated with large sky regions.