Original Input Image Research Articles

Enhancing Brain Tumor Diagnosis with L-Net: A Novel Deep Learning Approach for MRI Image Segmentation and Classification.

Background: Brain tumors are highly complex, making their detection and classification a significant challenge in modern medical diagnostics. The accurate segmentation and classification of brain tumors from MRI images are crucial for effective treatment planning. This study aims to develop an advanced neural network architecture that addresses these challenges. Methods: We propose L-net, a novel architecture combining U-net for tumor boundary segmentation and a convolutional neural network (CNN) for tumor classification. These two units are coupled such a way that the CNN classifies the MRI images based on the features extracted by the U-net while segmenting the tumor, instead of relying on the original input images. The model is trained on a dataset of 3064 high-resolution MRI images, encompassing gliomas, meningiomas, and pituitary tumors, ensuring robust performance across different tumor types. Results: L-net achieved a classification accuracy of up to 99.6%, surpassing existing models in both segmentation and classification tasks. The model demonstrated effectiveness even with lower image resolutions, making it suitable for diverse clinical settings. Conclusions: The proposed L-net model provides an accurate and unified approach to brain tumor segmentation and classification. Its enhanced performance contributes to more reliable and precise diagnosis, supporting early detection and treatment in clinical applications.

Unpaired data training enables super-resolution confocal microscopy from low-resolution acquisitions.

Supervised deep-learning models have enabled super-resolution imaging in several microscopic imaging modalities, increasing the spatial lateral bandwidth of the original input images beyond the diffraction limit. Despite their success, their practical application poses several challenges in terms of the amount of training data and its quality, requiring the experimental acquisition of large, paired databases to generate an accurate generalized model whose performance remains invariant to unseen data. Cycle-consistent generative adversarial networks (cycleGANs) are unsupervised models for image-to-image translation tasks that are trained on unpaired datasets. This paper introduces a cycleGAN framework specifically designed to increase the lateral resolution limit in confocal microscopy by training a cycleGAN model using low- and high-resolution unpaired confocal images of human glioblastoma cells. Training and testing performances of the cycleGAN model have been assessed by measuring specific metrics such as background standard deviation, peak-to-noise ratio, and a customized frequency content measure. Our cycleGAN model has been evaluated in terms of image fidelity and resolution improvement using a paired dataset, showing superior performance than other reported methods. This work highlights the efficacy and promise of cycleGAN models in tackling super-resolution microscopic imaging without paired training, paving the path for turning home-built low-resolution microscopic systems into low-cost super-resolution instruments by means of unsupervised deep learning.

DCPNet: Deformable Control Point Network for image enhancement

In this paper, we present a novel image enhancement network consisting of global and local color enhancement. The proposed network model is constructed using global transformation functions, which are formed by a set of piece-wise quadratic curves and a local color enhancement network based on the encoder–decoder network. To adaptively and dynamically control the ranges of each piece-wise curve, we introduce deformable control points (DCPs), which determine the overall structure of the global transformation functions. The parameters for piece-wise quadratic curve fitting and DCPs are estimated using the proposed DCP network (DCPNet). DCPNet processes a down-sampled image to derive the DCP parameters: The DCP offsets and the curve parameters. Then, we obtain a set of DCPs from the DCP offsets and connect each adjacent DCP pair by using the curve parameter to construct a global transformation function for each color channel. The original input images are then transformed based on the resulting transformation functions to obtain globally enhanced images. Finally, the intermediate image is fed into the local enhancement network, which has a U-Net architecture, to produce the spatially enhanced images. Extensive experimental results demonstrate the superiority of the proposed method over state-of-the-art methods in various image enhancement tasks, such as image retouching, tone-mapping, and underwater image enhancement.

Three-stage binarization of color document images based on discrete wavelet transform and generative adversarial networks

The efficient extraction of text information from the background in degraded color document images is an important challenge in the preservation of ancient manuscripts. The imperfect preservation of ancient manuscripts has led to different types of degradation over time, such as page yellowing, staining, and ink bleeding, seriously affecting the results of document image binarization. This work proposes an effective three-stage network method to image enhancement and binarization of degraded documents using generative adversarial networks (GANs). Specifically, in Stage-1, we first split the input images into multiple patches, and then split these patches into four single-channel patch images (gray, red, green, and blue). Then, three single-channel patch images (red, green, and blue) are processed by the discrete wavelet transform (DWT) with normalization. In Stage-2, we use four independent generators to separately train GAN models based on the four channels on the processed patch images to extract color foreground information. Finally, in Stage-3, we train two independent GAN models on the outputs of Stage-2 and the resized original input images (512 × 512) as the local and global predictions to obtain the final outputs. The experimental results show that the Avg-Score metrics of the proposed method are 77.64, 77.95, 79.05, 76.38, 75.34, and 77.00 on the (H)-DIBCO 2011, 2013, 2014, 2016, 2017, and 2018 datasets, which are at the state-of-the-art level. The implementation code for this work is available at https://github.com/abcpp12383/ThreeStageBinarization.

Low-Light Image Enhancement via Dual Information-Based Networks

Recently, deep-learning-based low-light image enhancement (LLIE) methods have made great progress. Benefiting from elaborately designed model architectures, these methods enjoy considerable performance gains. However, the generalizability of these methods may be weak, and they may suffer from the overfitting risk in the case of insufficient data as a result. At the same time, their complex model design brings serious computational burdens. To further improve performance, we exploit dual information, including spatial and channel (contextual) information, in the high-dimensional feature space. Specifically, we introduce customized spatial and channel blocks according to the feature difference of different layers. In shallow layers, the feature resolution is close to that of the original input image, and the spatial information is well preserved. Therefore, the spatial restoration block is designed for leveraging such precise spatial information to achieve better spatial restoration, e.g., revealing the textures and suppressing the noise in the dark. In deep layers, the features contain abundant contextual information, which is distributed in various channels. Hence, the channel interaction block is incorporated for better feature interaction, resulting in stronger model representation capability. Combining the U-Net-like model with the customized spatial and channel blocks makes up our method, which effectively utilizes dual information for image enhancement. Through extensive experiments, we demonstrate that our method, despite its simplicity of design, can provide advanced or competitive performance compared to some state-of-the-art deep learning- based methods.

Independently Trained Multi-Scale Registration Network Based on Image Pyramid.

Image registration is a fundamental task in various applications of medical image analysis and plays a crucial role in auxiliary diagnosis, treatment, and surgical navigation. However, cardiac image registration is challenging due to the large non-rigid deformation of the heart and the complex anatomical structure. To address this challenge, this paper proposes an independently trained multi-scale registration network based on an image pyramid. By down-sampling the original input image multiple times, we can construct image pyramid pairs, and design a multi-scale registration network using image pyramid pairs of different resolutions as the training set. Using image pairs of different resolutions, train each registration network independently to extract image features from the image pairs at different resolutions. During the testing stage, the large deformation registration is decomposed into a multi-scale registration process. The deformation fields of different resolutions are fused by a step-by-step deformation method, thereby addressing the challenge of directly handling large deformations. Experiments were conducted on the open cardiac dataset ACDC (Automated Cardiac Diagnosis Challenge); the proposed method achieved an average Dice score of 0.828 in the experimental results. Through comparative experiments, it has been demonstrated that the proposed method effectively addressed the challenge of heart image registration and achieved superior registration results for cardiac images.

DefenseFea: An Input Transformation Feature Searching Algorithm Based Latent Space for Adversarial Defense

Abstract Deep neural networks based image classification systems could suffer from adversarial attack algorithms, which generate input examples by adding deliberately crafted yet imperceptible noise to original input images. These crafted examples can fool systems and further threaten their security. In this paper, we propose to use latent space protect image classification. Specifically, we train a feature searching network to make up the difference between adversarial examples and clean examples with label guided loss function. We name it DefenseFea(input transformation based defense with label guided loss function), experimental result shows that DefenseFea can improve the rate of adversarial examples that achieved a success rate of about 99% on a specific set of 5000 images from ILSVRC 2012. This study plays a positive role in the further investigation of the relationship between adversarial examples and clean examples.

KernelFlexSR: a self-adaptive super-resolution algorithm with multi-path convolution and residual network for dynamic kernel enhancement

Machine learning-based image super-resolution (SR) has garnered increasing research interest in recent years. However, there are two issues that have not been adequately addressed. The first issue is that existing SR methods often overlook the importance of improving the quality of the training dataset, which is a crucial factor in determining SR performance, regardless of the training method employed. The second issue is that while some studies report high numerical metrics, the visual results remain unsatisfactory. To address the first problem, we propose a new image down-sampling method to obtain higher-quality training datasets. To tackle the second problem, we present a new image super-resolution model based on a large-size convolution kernel and a multi-path algorithm. Specifically, we use an adaptive large-size convolutional kernel to extract features from the image based on the size of the input image, and a residual network to generate a deeper model to retain more details of the original input image. Experimental results demonstrate that the proposed multilayer downsampling method (MDM) can significantly improve the visual quality compared to traditional downsampling methods. Moreover, our proposed method achieves the best peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) values compared to several typical SR algorithms. Furthermore, subjective evaluation by human observers reveals that our method retains more details of the original image and produces smoother high-resolution images. Our proposed method effectively addresses the two aforementioned issues, which leads to improved SR performance in terms of both quantitative and qualitative measures.

LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation

Medical image segmentation has always been a challenging task. This paper proposes a new LUneXt medical image segmentation model based on the characteristics analysis of medical image data sets and testing of different nonlinear activation units. Both normalized activations for the original negative input image activation have good optimization capabilities for the tokenization module parameters proposed in the original UneXt model. Using different activation coefficients for different foreground and background areas has achieved better results. The experimental results of this paper on the Breast Ultrasound Images (BUSI) data set reached an intersection over union (IoU) value of 62.64%, a Dice value of 76.12%, and a single inference speed of 807.57[Formula: see text]ms. The experimental IoU value of the International Skin Imaging Collaboration (ISIC 2018) data set reached 82.95%, and the Dice value reached 90.50%. The single inference speed reached 842.58[Formula: see text]ms. The LUneXt model is more robust than other models. While improving model performance, it does not introduce higher computational complexity and does not have a major impact on the processing speed of a single image.

Machine learning as a feedback tool for the gestalt principle of proximity

When teaching students about the underlying structure of images, the Gestalt Principles are frequently used. Students and teachers may not always agree on the Gestalt Principles that an image contains, making assessment activities focusing on this topic very subjective. When working with large class sizes or when using multiple markers, this subjectivity becomes more problematic. A DBSCAN-based feature generation process and a multi-layer perceptron-based model are suggested in this study for categorising images as exhibiting the principle of proximity. When classifying images as containing aspects of proximity, the model displays a high level of accuracy. The model was integrated as a component of a larger process that alters the original input images to highlight areas that demonstrate proximity. The use of these annotated images can aid in giving students feedback. However, as the model was trained on a synthetic data set, its accuracy on real-world student submissions is yet to be tested.

Image quality improvement in bowtie-filter-equipped cone-beam CT using a dual-domain neural network.

The bowtie-filter in cone-beam CT (CBCT) causes spatially nonuniform x-ray beam often leading to eclipse artifacts in the reconstructed image. The artifacts are further confounded by the patient scatter, which is therefore patient-dependent as well as system-specific. In this study, we propose a dual-domain network for reducing the bowtie-filter-induced artifacts in CBCT images. In the projection domain, the network compensates for the filter-induced beam-hardening that are highly related to the eclipse artifacts. The output of the projection-domain network was used for image reconstruction and the reconstructed images were fed into the image-domain network. In the image domain, the network further reduces the remaining cupping artifacts that are associated with the scatter. A single image-domain-only network was also implemented for comparison. The proposed approach successfully enhanced soft-tissue contrast with much-reduced image artifacts. In the numerical study, the proposed method decreased perceptual loss and root-mean-square-error (RMSE) of the images by 84.5% and 84.9%, respectively, and increased the structure similarity index measure (SSIM) by 0.26 compared to the original input images on average. In the experimental study, the proposed method decreased perceptual loss and RMSE of the images by 87.2% and 92.1%, respectively, and increased SSIM by 0.58 compared to the original input images on average. We have proposed a deep-learning-based dual-domain framework to reduce the bowtie-filter artifacts and to increase the soft-tissue contrast in CBCT images. The performance of the proposed method has been successfully demonstrated in both numerical and experimental studies.

A New Efficient Forgery Detection Method using Scaling, Binning, Noise Measuring Techniques and Artificial Intelligence (Ai)

In the market new updated editing tools and technologies are available to edit images and with help of these tools images are easily forged. In this research paper we proposed new forgery detection technique with estimation of noise on various scale of input image affect of noise in input image, frequency of images are also changed due to noise, noise signal correlated with original input images and in compressed images quantization level frequency also changed due to noise.We partition input image into M X N blocks, resized blocks are proceed further, image colors are also taken into consideration, each block noise value is evaluated at local level and global level. For each color channel of input image estimate local and global noise levels are estimated and compared using binning method. Also measured heat map of each block and each color channel of input image and all these values are stored in bins. Finally from all noise values calculate average mean value of noise, with these values decide whether input image is forgery or not, and performance of proposed method is compared with existing methods.

Data Class-Specific All-Optical Transformations and Encryption.

Diffractive optical networks provide rich opportunities for visual computing tasks since the spatial information of a scene can be directly accessed by a diffractive processor without requiring any digital pre-processing steps. Here we present data class-specific transformations all-optically performed between the input and output fields-of-view (FOVs) of a diffractive network. The visual information of the objects is encoded into the amplitude (A), phase (P), or intensity (I) of the optical field at the input, which is all-optically processed by a data class-specific diffractive network. At the output, an image sensor-array directly measures the transformed patterns, all-optically encrypted using the transformation matrices pre-assigned to different data classes, i.e., a separate matrix for each data class. The original input images can be recovered by applying the correct decryption key (the inverse transformation) corresponding to the matching data class, while applying any other key will lead to loss of information. The class-specificity of these all-optical diffractive transformations creates opportunities where different keys can be distributed to different users; each user can only decode the acquired images of only one data class, serving multiple users in an all-optically encrypted manner. We numerically demonstrated all-optical class-specific transformations covering A → A, I → I, and P → I transformations using various image datasets. We also experimentally validated the feasibility of this framework by fabricating class-specific I → I transformation diffractive networks and successfully tested them at different parts of the electromagnetic spectrum, i.e., 1550 nm and 0.75 mm wavelengths. Data class-specific all-optical transformations provide a fast and energy-efficient method for image and data encryption, enhancing data security and privacy. This article is protected by copyright. All rights reserved.

LEPF-Net: Light Enhancement Pixel Fusion Network for Underwater Image Enhancement

Underwater images often suffer from degradation due to scattering and absorption. With the development of artificial intelligence, fully supervised learning-based models have been widely adopted to solve this problem. However, the enhancement performance is susceptible to the quality of the reference images, which is more pronounced in underwater image enhancement tasks because the ground truths are not available. In this paper, we propose a light-enhanced pixel fusion network (LEPF-Net) to solve this problem. Specifically, we first introduce a novel light enhancement block (LEB) based on the residual block (RB) and the light enhancement curve (LE-Curve) to restore the cast color of the images. The RB is adopted to learn and obtain the feature maps from an original input image, and the LE-Curve is used to renovate the color cast of the learned images. To realize the superb detail of the repaired images, which is superior to the reference images, we develop a pixel fusion subnetwork (PF-SubNet) that adopts a pixel attention mechanism (PAM) to eliminate noise from the underwater image. The PAM adapts weight allocation to different levels of a feature map, which leads to an enhancement in the visibility of severely degraded areas. The experimental results show that the proposed LEPF-Net outperforms most of the existing underwater image enhancement methods. Furthermore, among the five classic no-reference image quality assessment (NRIQA) indicators, the enhanced images obtained by LEPF-Net are of higher quality than the ground truths from the UIEB dataset.

Iterative histogram equalization using discrete wavelet transform in low-dynamic range

A progressive histogram equalization (HE) and contrast enhancement method is proposed in the frequency domain. This method has an iterative structure based on the 6-sigma rule in the low-dynamic range using discrete wavelet transform. The proposed method determines an automatic threshold for attenuating the high-frequency amplitude of a signal in the discrete wavelet domain, based on the standard deviation of the absolute power of the high-frequency components in the signal band generated from the probability density function (PDF). Then an iterative quantization procedure is used where the PDF values in the frequency domain are randomly quantized. Besides, the maximum Bhattacharyya coefficient, which matches the original input image’s PDF, is used as a cost function to optimize the histogram smoothing output. For the random number generation, a 32-bit pseudorandom generator is employed to produce the same result for the output image at each runtime. The quantization factor parameter enables a controlled contrast enhancement rate to receive balanced equalization results in images. In the experimental studies, the proposed iterative HE method is compared with the well-known global, local, and brightness preserving algorithms. Experimental studies quantitatively and qualitatively display promising and encouraging results in terms of various state-of-the-art quality assessment metrics such as mean squared error, peak signal-to-noise ratio, structural similarity index measurement, contrast improvement index, absolute mean brightness error, and quality-aware relative contrast measure.

Domain generalization in nematode classification

Nematode images captured by different microscopes may appear differently in terms of image color and image quality, resulting in these images laying in different learning domains. This can negatively impact nematode classification via deep learning. In this paper, we propose a local structure invariance guided (LSIG) domain generalization approach to enhance the model generalization of nematode local regions in unseen domains. First, a style transfer method is introduced to synthesize new domain image samples from the source domain. Unlike in the original input images, the color information of the synthetic images is changed, but their structural information is retained. Then, a metric learning strategy is designed to determine the cross-domain invariant structural representation between the source and new domains by pairwise learning. Each class is then effectively clustered, and a better decision boundary is determined to improve the model generalization. Overall, we demonstrate the effectiveness and robustness of the method on binary-class and multi-class classification tasks on diverse nematode datasets.

Enhancing the Breast Histopathology Image Analysis for Cancer Detection Using Variational Autoencoder.

A breast tissue biopsy is performed to identify the nature of a tumour, as it can be either cancerous or benign. The first implementations involved the use of machine learning algorithms. Random Forest and Support Vector Machine (SVM) were used to classify the input histopathological images into whether they were cancerous or non-cancerous. The implementations continued to provide promising results, and then Artificial Neural Networks (ANNs) were applied for this purpose. We propose an approach for reconstructing the images using a Variational Autoencoder (VAE) and the Denoising Variational Autoencoder (DVAE) and then use a Convolutional Neural Network (CNN) model. Afterwards, we predicted whether the input image was cancerous or non-cancerous. Our implementation provides predictions with 73% accuracy, which is greater than the results produced by our custom-built CNN on our dataset. The proposed architecture will prove to be a new field of research and a new area to be explored in the field of computer vision using CNN and Generative Modelling since it incorporates reconstructions of the original input images and provides predictions on them thereafter.

UIDEF: A real-world underwater image dataset and a color-contrast complementary image enhancement framework

Numerous underwater image enhancement methods have been proposed in recent years. However, these methods are mainly evaluated using synthetic datasets with similar degradation or real-world datasets with insufficient images. A benchmark dataset containing various degraded situations and real-world underwater images is needed to evaluate the performance of these methods. In this paper, we propose a Real-world Underwater Image Dataset (UIDEF), which consists of seven categories and eleven subcategories with 9200 real-world underwater images. These categories roughly cover the multiple degradation types and different shooting perspectives of common underwater imagery. Using this dataset, we conduct a qualitative and quantitative empirical comparison of eight state-of-the-art underwater image enhancement methods to evaluate their effectiveness and robustness. Considering that these methods cannot handle both color restoration and contrast enhancement of the underwater degraded images well, we present a color-contrast complementary image enhancement framework that consists of the adaptive color perception balance and multi-scale weighted fusion. The former procedure is essential to remove the color cast of original input images, while the latter defines four attentive weight maps for making the enhanced output images present a more comfortable visual perception. Extensive experiments have validated that our framework achieves relatively satisfactory performance in most cases. In addition, we demonstrate an additional application of UIDEF in reconstructing a wider-field underwater image based on multiple images with overlapping regions. The dataset is available at https://github.com/LaibinChang/UIDEF.git.

Fault diagnosis for overcharge and undercharge conditions in refrigeration systems using infrared thermal images

With the increase in industrialization and the number of people and buildings, the need for refrigeration systems has also increased. Maintenance of these systems, malfunctions and their late detection cause time and cost problems. Therefore, in this study, a machine learning application is recommended to diagnose the refrigerant undercharge and refrigerant overcharge faults that may occur in the refrigeration system by using infrared images. Firstly, infrared images obtained from normal charge, undercharge and overcharge situations in the refrigeration system, are passed through the two-dimensional discrete wavelet transform (2D-DWT) and the images are decomposed. Then, statistical texture features from the original input images are obtained by separating infrared images. The dimensionality of the extracted features is reduced using the principal component analysis (PCA) and the ReliefF (RF) algorithm. Finally, these selected features are applied to the K nearest neighbor (k-NN), naive Bayes algorithm (NBA), decision tree (DT), and cascade forward neural network (CFNN) classifiers. It has been found that RF-based feature selection is useful in obtaining the optimal feature vector. The classification results revealed that CFNN outperforms k-NN, NBA, and DT. Compared to traditional electrical measurements and fault detection methods, it has been shown that the recommended system is feasible due to its features such as ease of use, remote measurement, and self-adaptive recognition.

Novel Contiguous Cross Propagation Neural Network Built CAD for Lung Cancer

The present progress of visual-based detection of the diseased area of a malady plays an essential part in the medical field. In that case, the image processing is performed to improve the image data, wherein it inhibits unintended distortion of image features or it enhances further processing in various applications and fields. This helps to show better results especially for diagnosing diseases. Of late the early prediction of cancer is necessary to prevent disease-causing problems. This work is proposed to identify lung cancer using lung computed tomography (CT) scan images. It helps to identify cancer cells’ affected areas. In the present work, the original input image from Lung Image Database Consortium (LIDC) typically suffers from noise problems. To overcome this, the Gabor filter used for image processing is highly enhanced. In the next stage, the Spherical Iterative Refinement Clustering (SIRC) algorithm identifies cancer-suspected areas on the CT scan image. This approach can help radiologists and medical experts recognize cancer diseases and syndromes so that serious progress can be avoided in the early stages. These new methods help to remove unwanted portions of the CT image and better utilization the image. The subspace extraction of features approach is beneficial for evaluating lung cancer. This paper introduces a novel approach called Contiguous Cross Propagation Neural Network that tends to locate regions afflicted by lung cancer using CT scan pictures (CCPNN). By using the feature values from the fourth step of the procedure, the proposed CCPNN tends to categorize the lesion in the lung nodular site. The efficiency of the suggested CCPNN approach is evaluated using classification metrics such as recall (%), precision (%), F-measure (percent), and accuracy (%). Finally, the incorrect classification ratios are determined to compare the trained networks’ effectiveness, through these parameters of CCPNN, it obtains the outstanding performance of 98.06% and it has provided the lowest false ratio of 1.8%.