Dense Block Research Articles

Super-Resolution Reconstruction of CT Images Based on Multi-scale Information Fused Generative Adversarial Networks.

The popularization and widespread use of computed tomography (CT) in the field of medicine evocated public attention to the potential radiation exposure endured by patients. Reducing the radiation dose may lead to scattering noise and low resolution, which can adversely affect the radiologists' judgment. Hence, this paper introduces a new network called PANet-UP-ESRGAN (PAUP-ESRGAN), specifically designed to obtain low-dose CT (LDCT) images with high peak signal-to-noise ratio (PSNR) and high resolution (HR). The model was trained on synthetic medical image data based on a Generative Adversarial Network (GAN). A degradation modeling process was introduced to accurately represent realistic degradation complexities. To reconstruct image edge textures, a pyramidal attention model call PANet was added before the middle of the multiple residual dense blocks (MRDB) in the generator to focus on high-frequency image information. The U-Net discriminator with spectral normalization was also designed to improve its efficiency and stabilize the training dynamics. The proposed PAUP-ESRGAN model was evaluated on the abdomen and lung image datasets, which demonstrated a significant improvement in terms of robustness of model and LDCT image detail reconstruction, compared to the latest real-esrgan network. Results showed that the mean PSNR increated by 19.1%, 25.05%, and 21.25%, the mean SSIM increated by 0.4%, 0.4%, and 0.4%, and the mean NRMSE decreated by 0.25%, 0.25%, and 0.35% at 2[Formula: see text], 4[Formula: see text], and 8[Formula: see text] super-resolution scales, respectively. Experimental results demonstrate that our method outperforms the state-of-the-art super-resolution methods on restoring CT images with respect to peak signal-to-noise ratio (PSNR), structural similarity (SSIM) and normalized root-mean-square error (NRMSE) indices.

An image deblurring method using improved U-Net model based on multilayer fusion and attention mechanism

The investigation of image deblurring techniques in dynamic scenes represents a prominent area of research. Recently, deep learning technology has gained extensive traction within the field of image deblurring methodologies. However, such methods often suffer from limited inherent interconnections across various hierarchical levels, resulting in inadequate receptive fields and suboptimal deblurring outcomes. In U-Net, a more adaptable approach is employed, integrating diverse levels of features effectively. Such design not only significantly reduces the number of parameters but also maintains an acceptable accuracy range. Based on such advantages, an improved U-Net model for enhancing the image deblurring effect was proposed in the present study. Firstly, the model structure was designed, incorporating two key components: the MLFF (multilayer feature fusion) module and the DMRFAB (dense multi-receptive field attention block). The aim of these modules is to improve the feature extraction ability. The MLFF module facilitates the integration of feature information across various layers, while the DMRFAB module, enriched with an attention mechanism, extracts crucial and intricate image details, thereby enhancing the overall information extraction process. Finally, in combination with fast Fourier transform, the FRLF (Frequency Reconstruction Loss Function) was proposed. The FRLF obtains the frequency value of the image by reducing the frequency difference. The present experiment results reveal that the proposed method exhibited higher-quality visual effects. Specifically, for the GoPro dataset, the PSNR (peak signal-to-noise ratio) reached 31.53, while the SSIM (structural similarity index) attained a value of 0.948. Additionally, for the Real Blur dataset, the PSNR achieved 31.32, accompanied by an SSIM score of 0.934.

SNO-DCA: A model for predicting S-nitrosylation sites based on densely connected convolutional networks and attention mechanism

Protein S-nitrosylation is a reversible oxidative reduction post-translational modification that is widely present in the biological community. S-nitrosylation can regulate protein function and is closely associated with a variety of diseases, thus identifying S-nitrosylation sites are crucial for revealing the function of proteins and related drug discovery. Traditional experimental methods are time-consuming and expensive; therefore, it is necessary to explore more efficient computational methods. Deep learning algorithms perform well in the field of bioinformatics sites prediction, and many studies show that they outperform existing machine learning algorithms. In this work, we proposed a deep learning algorithm-based predictor SNO-DCA for distinguishing between S-nitrosylated and non-S-nitrosylated sequences. First, one-hot encoding of protein sequences was performed. Second, the dense convolutional blocks were used to capture feature information, and an attention module was added to weigh different features to improve the prediction ability of the model. The 10-fold cross-validation and independent testing experimental results show that our SNO-DCA model outperforms existing S-nitrosylation sites prediction models under imbalanced data. In this paper, a web server prediction website: https://sno.cangmang.xyz/SNO-DCA/was established to provide an online prediction service for users. SNO-DCA can be available at https://github.com/peanono/SNO-DCA.

An inverse halftoning method based on supervised deep convolutional neural network

AbstractInverse halftoning is a technology that converts a binary image into a continuous tone image. Due to the wide application of inverse halftoning, many scholars have proposed several deep convolutional neural networks (DCNN) to optimize their performance. According to the observation, there is still room for improvement in content generation and detail recovery of the inverse halftone images generated by using the existing methods. Therefore, an inverse halftoning method based on supervised DCNN is proposed in this paper. The method consists of two parts: the multi‐level feature extraction model uses the down‐sampling to extract the features from the halftone image and remove the halftone noise dots on flat areas, which is implemented by four convolutional layers; the image reconstruction model uses up‐sampling to reconstruct image information, which is realized by four convolutional layers and two dense residual blocks. At the same time, in order to further recover the details, the down‐sampling feature maps and up‐sampling feature maps of the same size are concatenated by addition layers. Experimental results show that compared with other methods, the inverse halftone images obtained by the proposed network have better results in both subjective and objective evaluations.

Satellite image fusion using cyclic spatio-spectral GAN model

ABSTRACT Preserving spectral and spatial information in the satellite image fusion problem is one of the essential challenges. This paper presents a GAN-based method for fusing panchromatic and multispectral satellite images. The proposed method utilises the idea of Cycle-GAN with two generators, for spectral and spatial information preservation, based on residual in residual dense block super-resolution architecture. First, generator-1 translates panchromatic and multispectral satellite images to a high-resolution fused image and then preserves details by utilising generator-2. The goal is to reach the spatial and spectral details of panchromatic and multispectral satellite images, respectively. Two discriminators are employed, one for spectral and the other for spatial transformations, and weighted L1 loss is used for both as cycle losses. By leveraging the unique capabilities of the two generators, the proposed method achieves high-quality image fusion results with improved spectral and spatial resolutions that are evaluated over two different datasets. The experimental results demonstrate the effectiveness of the proposed method, with enhancements ranging from approximately 2% to 30% and 0.5% to 35% in WorldView2 and GeoEye-1 images, respectively, in five metrics including PSNR. Moreover, we show significant enhancements of approximately 7% to 50% and 22% to 29% in the same datasets in metrics like SAM and ERGAS.

Densely connected convolutional networks for ultrasound image based lesion segmentation

Delineating lesion boundaries play a central role in diagnosing thyroid and breast cancers, making related therapy plans and evaluating therapeutic effects. However, it is often time-consuming and error-prone with limited reproducibility to manually annotate low-quality ultrasound (US) images, given high speckle noises, heterogeneous appearances, ambiguous boundaries etc., especially for nodular lesions with huge intra-class variance. It is hence appreciative but challenging for accurate lesion segmentations from US images in clinical practices. In this study, we propose a new densely connected convolutional network (called MDenseNet) architecture to automatically segment nodular lesions from 2D US images, which is first pre-trained over ImageNet database (called PMDenseNet) and then retrained upon the given US image datasets. Moreover, we also designed a deep MDenseNet with pre-training strategy (PDMDenseNet) for segmentation of thyroid and breast nodules by adding a dense block to increase the depth of our MDenseNet. Extensive experiments demonstrate that the proposed MDenseNet-based method can accurately extract multiple nodular lesions, with even complex shapes, from input thyroid and breast US images. Moreover, additional experiments show that the introduced MDenseNet-based method also outperforms three state-of-the-art convolutional neural networks in terms of accuracy and reproducibility. Meanwhile, promising results in nodular lesion segmentation from thyroid and breast US images illustrate its great potential in many other clinical segmentation tasks.

Perception-oriented generative adversarial network for retinal fundus image super-resolution

Retinal fundus imaging is a crucial diagnostic tool in ophthalmology, enabling the early detection and monitoring of various ocular diseases. However, capturing high-resolution fundus images often presents challenges due to factors such as defocusing and diffraction in the digital imaging process, limited shutter speed, sensor unit density, and random noise in the image sensor or during image transmission. Super-resolution techniques offer a promising solution to overcome these limitations and enhance the visual details in retinal fundus images. Since the retina has rich texture details, the super-resolution images often introduce artifacts into texture details and lose some fine retinal vessel structures. To improve the perceptual quality of the retinal fundus image, a generative adversarial network that consists of a generator and a discriminator is proposed. The proposed generator mainly comprises 23 multi-scale feature extraction blocks, an image segmentation network, and 23 residual-in-residual dense blocks. These components are employed to extract features at different scales, acquire the retinal vessel grayscale image, and extract retinal vascular features, respectively. The generator has two branches that are mainly responsible for extracting global features and vascular features, respectively. The extracted features from the two branches are fused to better restore the super-resolution image. The proposed generator can restore more details and more accurate fine vessel structures in retinal images. The improved discriminator is proposed by introducing our designed attention modules to help the generator yield clearer super-resolution images. Additionally, an artifact loss function is also introduced to enhance the generative adversarial network, enabling more accurate measurement of the disparity between the high-resolution image and the restored image. Experimental results show that the generated images obtained by our proposed method have a better perceptual quality than the state-of-the-art image super-resolution methods.

ViT-DexiNet: a vision transformer-based edge detection operator for small object detection in SAR images

ABSTRACTThis paper introduces a novel edge detection operator called ‘vision transformer-based Dexinet (ViT-DexiNet)’ to address the challenges of detecting small objects in synthetic aperture radar (SAR) images. SAR images are typically impacted by strong multiplicative noise, making edge detection difficult. Existing traditional methods have limited spectral data preservation capabilities and often result in a loss of clarity and integrity of salient features in SAR images. The proposed ViT-DexiNet operator employs a series of interconnected layers to extract and refine salient edge features from SAR images. It utilizes a vision transformer self-attention layer to capture the pattern and structural details of image features crucial for determining edges. The extracted feature maps are then processed by the DexiNet architecture, which consists of a dense block, transfer block, and upsampling network. This architecture helps preserve edge information at different scales in deeper layers. The series of layered blocks generate edge maps, which are concatenated and averaged through smoothing to remove noise and enhance edge details in SAR images, resulting in a final high-quality edge map. To evaluate the proposed ViT-DexiNet method, both qualitative and quantitative analyses are conducted using standard edge detection operators such as Canny and Sobel. The empirical results demonstrate that the ViT-DexiNet surpasses baseline edge detection operators. The achieved values of proposed edge detection operator are 97.92%, 97.72%, 97.64% and 97.41%, respectively for the metrics accuracy, precision, recall and f1-score. The ViT-DexiNet offers high-quality edge maps, simplifying the interpretation of data for small object detection. Overall, the ViT-DexiNet method shows promise in overcoming the limitations of traditional approaches and improving the detection of edges in SAR images.

Novel fast preparation of MoAlB phase cermet with elevated temperature oxidation resistance

Owing to the small production scale, high energy consumption and low degree of the current sintering method, the facile and fast method to prepare tailored MoAlB phase cermets is urgent. Hence, the high purity, dense and large-size MoAlB phase cermet blocks are fabricated by a novel direct current (DC) fast hot pressing (FHP) sintering technology starting from Mo/Al/B powders with a molar ratio of 1:1.3:1. The phase composition and the microstructure of MoAlB phase cermets under different sintering conditions are characterized with X-ray diffraction and scanning electron microscopy. The results indicate that the as-prepared MoAlB samples are almost single-phase. MoAlB phase cermet has good oxidation resistance, forming a thin layer with Al2O3, B2O3 and MoO3 at 1000 °C.

A novel dual-stage progressive enhancement network for single image deraining

The dense rain accumulation in heavy rain can significantly wash out images and thus destroy the background details of images. Although existing deep rain removal models lead to improved performance for heavy rain removal, we find that most of them ignore the detail reconstruction accuracy of rain-free images. In this paper, we propose a dual-stage progressive enhancement network (DPENet-v2) to achieve effective deraining with structure-accurate rain-free images. Three main modules are included in our framework, namely a rain streaks removal network (R2Net), a details reconstruction network (DRNet) and a cross-stage feature interaction module (CFIM). The former aims to achieve accurate rain removal, and the latter is designed to recover the details of rain-free images. We introduce two main strategies within our networks to achieve trade-off between the effectiveness of deraining and the detail restoration of rain-free images. Firstly, a dilated dense residual block (DDRB) within the rain streaks removal network is presented to aggregate high/low level features of heavy rain. Secondly, an enhanced residual pixel-wise attention block (ERPAB) within the details reconstruction network is designed for context information aggregation. Meanwhile, CFIM learns the long-range dependencies and achieves cross-stage information communication. We also propose a comprehensive loss function to highlight the marginal and regional accuracy of rain-free images. Extensive experiments on benchmark public datasets show both efficiency and effectiveness of the proposed method in achieving structure-preserving rain-free images for heavy rain removal.

Synthetic Data Generation Based on RDB-CycleGAN for Industrial Object Detection

Deep learning-based methods have demonstrated remarkable success in object detection tasks when abundant training data are available. However, in the industrial domain, acquiring a sufficient amount of training data has been a challenge. Currently, many synthetic datasets are created using 3D modeling software, which can simulate real-world scenarios and objects but often cannot achieve complete accuracy and realism. In this paper, we propose a synthetic data generation framework for industrial object detection tasks based on image-to-image translation. To address the issue of low image quality that can arise during the image translation process, we have replaced the original feature extraction module with the Residual Dense Block (RDB) module. We employ the RDB-CycleGAN network to transform CAD models into realistic images. Additionally, we have introduced the SSIM loss function to strengthen the network constraints of the generator and conducted a quantitative analysis of the improved RDB-CycleGAN-generated synthetic data. To evaluate the effectiveness of our proposed method, the synthetic data we generate effectively enhance the performance of object detection algorithms on real images. Compared to using CAD models directly, synthetic data adapt better to real-world scenarios and improve the model’s generalization ability.

GAN-based super-resolution for confocal superficial eyelid imaging: Real-time, domain generalization, and noise robustness

Objective:Infection by Demodex mites is a significant factor in causing blepharitis. The collected superficial eyelid tissue images currently have low resolution and strong noise. High-quality reconstruction of low-quality images is needed to assist ophthalmologists in accurately and quickly diagnosing mite infestations. Methods:We propose a novel GAN-based super-resolution reconstruction model. We utilize Residual-in-Residual Dense Blocks (RRDB) as the fundamental building units and employ high-order degradation to simulate real-world degradation. Additionally, we introduce a new attention mechanism, All Attention Mechanism (AAM), which effectively captures and utilizes crucial image features, expanding the scope of utilized pixels. This significantly enhances the quality of reconstructed images. Results:Compared to existing super-resolution reconstruction models, our model demonstrates the strongest generalization ability and robustness. On the CIDMG and FMD datasets, we achieve the lowest NIQE and the highest SSIM values. Our model outperforms the SOTA SwinIR model, reducing reconstruction time by 76% and improving the image quality by 6.6% for low-noise images and 15.3% for high-noise images. Conclusion:By combining GAN with AAM, we have significantly improved the model’s generalization ability, robustness, and real-time performance. Significance:This is the first time that domain generalization has been used for reconstructing superficial eyelid tissue images. The stable and fast reconstruction performance reduces the impact of noise and low resolution on image quality, significantly reducing examination time and improving diagnostic accuracy. This greatly enhances the clinical utility of confocal microscopy for eyelid demodex examination.

Multikernel positional embedding convolutional neural network for photoacoustic reconstruction with sparse data.

Photoacoustic imaging (PAI) is an emerging noninvasive imaging modality that merges the high contrast of optical imaging with the high resolution of ultrasonic imaging. Low-quality photoacoustic reconstruction with sparse data due to sparse spatial sampling and limited view detection is a major obstacle to the popularization of PAI for medical applications. Deep learning has been considered as the best solution to this problem in the past decade. In this paper, we propose what we believe to be a novel architecture, named DPM-UNet, which consists of the U-Net backbone with additional position embedding block and two multi-kernel-size convolution blocks, a dilated dense block and dilated multi-kernel-size convolution block. Our method was experimentally validated with both simulated data and in vivo data, achieving a SSIM of 0.9824 and a PSNR of 33.2744dB. Furthermore, the reconstructed images of our proposed method were compared with those obtained by other advanced methods. The results have shown that our proposed DPM-UNet has a great advantage in PAI over other methods with respect to the imaging effect and memory consumption.

Fine Perceptive GANs for Brain MR Image Super-Resolution in Wavelet Domain.

Magnetic resonance (MR) imaging plays an important role in clinical and brain exploration. However, limited by factors such as imaging hardware, scanning time, and cost, it is challenging to acquire high-resolution MR images clinically. In this article, fine perceptive generative adversarial networks (FP-GANs) are proposed to produce super-resolution (SR) MR images from the low-resolution counterparts. By adopting the divide-and-conquer scheme, FP-GANs are designed to deal with the low-frequency (LF) and high-frequency (HF) components of MR images separately and parallelly. Specifically, FP-GANs first decompose an MR image into LF global approximation and HF anatomical texture subbands in the wavelet domain. Then, each subband generative adversarial network (GAN) simultaneously concentrates on super-resolving the corresponding subband image. In generator, multiple residual-in-residual dense blocks are introduced for better feature extraction. In addition, the texture-enhancing module is designed to trade off the weight between global topology and detailed textures. Finally, the reconstruction of the whole image is considered by integrating inverse discrete wavelet transformation in FP-GANs. Comprehensive experiments on the MultiRes_7T and ADNI datasets demonstrate that the proposed model achieves finer structure recovery and outperforms the competing methods quantitatively and qualitatively. Moreover, FP-GANs further show the value by applying the SR results in classification tasks.

Anti-Rounding Image Steganography With Separable Fine-Tuned Network

Image steganographic methods based on encoder-decoder model with end-to-end network architecture recently have been proposed. However, in steganographic applications, the feature map (called stego matrix) generated by the encoder needs to be rounded as a real stego image for the receiver. The loss of precision by rounding stego matrix leads to the decline in the accuracy of extracted secret messages. The challenge of using end-to-end network to preserve robustness against rounding operation is that it is non-differentiable. In this paper, we propose an anti-rounding image steganography method with separable fine-tuning network architecture which includes the joint training stage (JT-stage) and the separable fine-tuning stage (SF-stage). Firstly, in JT-stage, an embedded generator and a stego matrix extractor are jointly learned without rounding operation. Utilizing concatenation in embedded generator can realistically fuse cover image and secret messages. And the multi-scale fusion block and residual dense block in stego matrix extractor can make secret messages more correctly decoded. Moreover, the discriminator is constructed by generative adversarial nets (GAN) in JT-stage to effectively improve the authenticity and steganalysis security. Then, in SF-stage, the embedded generator is frozen, and the stego matrix is obtained and rounded as a stego image. A stego image extractor is constructed by fine-tuning the layers of the stego matrix extractor to improve the accuracy of message extraction. As the loss will not backpropagate in the embedded generator, the non-differentiability of rounding operation can be offset. Experiments show that the proposed separation fine-tuning network is robust to rounding operation, and effectively reduces the degradation of the image quality and steganalysis performance.

Lightweight 3D Dense Autoencoder Network for Hyperspectral Remote Sensing Image Classification.

The lack of labeled training samples restricts the improvement of Hyperspectral Remote Sensing Image (HRSI) classification accuracy based on deep learning methods. In order to improve the HRSI classification accuracy when there are few training samples, a Lightweight 3D Dense Autoencoder Network (L3DDAN) is proposed. Structurally, the L3DDAN is designed as a stacked autoencoder which consists of an encoder and a decoder. The encoder is a hybrid combination of 3D convolutional operations and 3D dense block for extracting deep features from raw data. The decoder composed of 3D deconvolution operations is designed to reconstruct data. The L3DDAN is trained by unsupervised learning without labeled samples and supervised learning with a small number of labeled samples, successively. The network composed of the fine-tuned encoder and trained classifier is used for classification tasks. The extensive comparative experiments on three benchmark HRSI datasets demonstrate that the proposed framework with fewer trainable parameters can maintain superior performance to the other eight state-of-the-art algorithms when there are only a few training samples. The proposed L3DDAN can be applied to HRSI classification tasks, such as vegetation classification. Future work mainly focuses on training time reduction and applications on more real-world datasets.

DESPP-DETR: A Dense Connection Efficient Spatial Pooling DEtection TRansformer for Vehicle Detection

Real-time vehicle detection is a challenging and vital task in intelligent transportation systems. The key requirements for a vehicle detection model are speed and accuracy. However, existing real-time vehicle detection models often sacrifice one of these qualities in favor of the other. This trade-off makes them unfit for real-time deployment, where both speed and accuracy are equally important. Additionally, occlusion, which refers to the obstruction or partial covering of vehicles, further complicates detection and affects the system’s accuracy. In this study, we propose DESPP-DETR, a one-stage detection network for real-time vehicle detection. It is based on bipartite matching and a transformer encoder-decoder architecture, with the addition of a dense connection block and enhanced spatial pyramid pooling. The presence of dense connection block strengthens feature extraction. The enhanced spatial pyramid pooling eliminates the fixed-size constraint and increases the network’s learning capacity. When compared to existing models, DESPP-DETR achieves greater accuracy in real-time vehicle detection. On the MS COCO 2017 dataset, the proposed model achieves an improved mean average precision (mAP) of 75.53%, making it a promising solution for intelligent transportation systems.

DeepHipp: accurate segmentation of hippocampus using 3D dense-block based on attention mechanism

BackgroundThe hippocampus is a key area of the brain responsible for learning, memory, and other abilities. Accurately segmenting the hippocampus and precisely calculating the volume of the hippocampus is of great significance for predicting Alzheimer’s disease and amnesia. Most of the segmentation algorithms currently involved are based on templates, such as the more popular FreeSufer.MethodsThis study proposes Deephipp, a deep learning network based on a 3D dense block using an attention mechanism for accurate segmentation of the hippocampus. DeepHipp is based on the following novelties: (i) DeepHipp adopts powerful data augmentation schemes to enhance the segmentation ability. (ii) DeepHipp is designed to incorporate 3D dense-block to capture multiple-scale features of the hippocampus. (iii) DeepHipp creatively uses the attention mechanism in the field of hippocampal image segmentation, extracting useful hippocampus information in a massive feature map, and improving the accuracy and sensitivity of the model.ConclusionsWe describe the illustrative results and show extensive qualitative and quantitative comparisons with other methods. Our achievement demonstrates that the accuracy of DeepHipp can reach 83.63%, which is superior to most existing methods in terms of accuracy and efficiency of hippocampus segmentation. It is noticeable that deep learning can potentially lead to an effective segmentation of medical images.

MixNet-LD: An Automated Classification System for Multiple Lung Diseases Using Modified MixNet Model.

The lungs are critical components of the respiratory system because they allow for the exchange of oxygen and carbon dioxide within our bodies. However, a variety of conditions can affect the lungs, resulting in serious health consequences. Lung disease treatment aims to control its severity, which is usually irrevocable. The fundamental objective of this endeavor is to build a consistent and automated approach for establishing the intensity of lung illness. This paper describes MixNet-LD, a unique automated approach aimed at identifying and categorizing the severity of lung illnesses using an upgraded pre-trained MixNet model. One of the first steps in developing the MixNet-LD system was to build a pre-processing strategy that uses Grad-Cam to decrease noise, highlight irregularities, and eventually improve the classification performance of lung illnesses. Data augmentation strategies were used to rectify the dataset's unbalanced distribution of classes and prevent overfitting. Furthermore, dense blocks were used to improve classification outcomes across the four severity categories of lung disorders. In practice, the MixNet-LD model achieves cutting-edge performance while maintaining model size and manageable complexity. The proposed approach was tested using a variety of datasets gathered from credible internet sources as well as a novel private dataset known as Pak-Lungs. A pre-trained model was used on the dataset to obtain important characteristics from lung disease images. The pictures were then categorized into categories such as normal, COVID-19, pneumonia, tuberculosis, and lung cancer using a linear layer of the SVM classifier with a linear activation function. The MixNet-LD system underwent testing in four distinct tests and achieved a remarkable accuracy of 98.5% on the difficult lung disease dataset. The acquired findings and comparisons demonstrate the MixNet-LD system's improved performance and learning capabilities. These findings show that the proposed approach may effectively increase the accuracy of classification models in medicinal image investigations. This research helps to develop new strategies for effective medical image processing in clinical settings.

DRGAN: Dense Residual Generative Adversarial Network for Image Enhancement in an Underwater Autonomous Driving Device.

Underwater autonomous driving devices, such as autonomous underwater vehicles (AUVs), rely on visual sensors, but visual images tend to produce color aberrations and a high turbidity due to the scattering and absorption of underwater light. To address these issues, we propose the Dense Residual Generative Adversarial Network (DRGAN) for underwater image enhancement. Firstly, we adopt a multi-scale feature extraction module to obtain a range of information and increase the receptive field. Secondly, a dense residual block is proposed, to realize the interaction of image features and ensure stable connections in the feature information. Multiple dense residual modules are connected from beginning to end to form a cyclic dense residual network, producing a clear image. Finally, the stability of the network is improved via adjustment to the training with multiple loss functions. Experiments were conducted using the RUIE and Underwater ImageNet datasets. The experimental results show that our proposed DRGAN can remove high turbidity from underwater images and achieve color equalization better than other methods.