Search Results (129)

Brain tumor detection and categorization of its subtypes are essential for early diagnosis and improving patient outcomes. This research presents a cutting-edge approach that employs advanced data augmentation and deep learning methodologies for brain tumor classification. For this work, a dataset of 6982 MRI images from the IEEE Data Port was considered, in which a total of 5712 images of four classes (1321 glioma, 1339 meningioma, 1595 no tumor, and 1457 pituitary) were used in the training set and a total of 1270 images of the same four classes were used in the testing set. A Wasserstein Generative Adversarial Network was implemented to generate synthetic images to address class imbalance, resulting in a balanced and consistent dataset. A comparison was conducted between various data augmentation metholodogies demonstrating that Wasserstein Generative Adversarial Network-augmented results perform excellently over traditional augmentation (such as rotation, shift, zoom, etc.) and no augmentation. Additionally, a Gaussian filter and normalization were applied during preprocessing to reduce noise, highlighting its superior accuracy and edge preservation by comparing its performance to Median and Bilateral filters. The classifier model combines parallel feature extraction from modified InceptionV3 and VGG19 followed by custom attention mechanisms for effectively capturing the characteristics of each tumor type. The model was trained for 64 epochs using model checkpoints to save the best-performing model based on validation accuracy and learning rate adjustments. The model achieved a 99.61% accuracy rate on the testing set, with precision, recall, AUC, and loss of 0.9960, 0.9960, 0.0153, and 0.9999, respectively. The proposed architecture’s explainability has been enhanced by t-SNE plots, which show unique tumor clusters, and Grad-CAM representations, which highlight crucial areas in MRI scans. This research showcases an explainable and robust approach for correctly classifying four brain tumor types, combining WGAN-augmented data with advanced deep learning models in feature extraction. The framework effectively manages class imbalance and integrates a custom attention mechanism, outperforming other models, thereby improving diagnostic accuracy and reliability in clinical settings. Full article

In the field of dentistry, the presence of dental calculus is a commonly encountered issue. If not addressed promptly, it has the potential to lead to gum inflammation and eventual tooth loss. Bitewing (BW) images play a crucial role by providing a comprehensive visual representation of the tooth structure, allowing dentists to examine hard-to-reach areas with precision during clinical assessments. This visual aid significantly aids in the early detection of calculus, facilitating timely interventions and improving overall outcomes for patients. This study introduces a system designed for the detection of dental calculus in BW images, leveraging the power of YOLOv8 to identify individual teeth accurately. This system boasts an impressive precision rate of 97.48%, a recall (sensitivity) of 96.81%, and a specificity rate of 98.25%. Furthermore, this study introduces a novel approach to enhancing interdental edges through an advanced image-enhancement algorithm. This algorithm combines the use of a median filter and bilateral filter to refine the accuracy of convolutional neural networks in classifying dental calculus. Before image enhancement, the accuracy achieved using GoogLeNet stands at 75.00%, which significantly improves to 96.11% post-enhancement. These results hold the potential for streamlining dental consultations, enhancing the overall efficiency of dental services. Full article

Transient Luminous Events (TLEs) are short-lived, upper-atmospheric optical phenomena associated with thunderstorms. Their rapid and random occurrence makes manual classification laborious and time-consuming. This study presents an effective approach to automating the classification of TLEs using state-of-the-art Convolutional Neural Networks (CNNs) and a Vision Transformer (ViT). The ViT architecture and four different CNN architectures, namely, ResNet50, ResNet18, GoogLeNet, and SqueezeNet, are employed and their performance is evaluated based on their accuracy and execution time. The models are trained on a dataset that was augmented using rotation, translation, and flipping techniques to increase its size and diversity. Additionally, the images are preprocessed using bilateral filtering to enhance their quality. The results show high classification accuracy across all models, with ResNet50 achieving the highest accuracy. However, a trade-off is observed between accuracy and execution time, which should be considered based on the specific requirements of the task. This study demonstrates the feasibility and effectiveness of using transfer learning and pre-trained CNNs for the automated classification of TLEs. Full article

The performance of the tire has a very important impact on the safe driving of the car, and in the actual use of the tire, due to complex road conditions or use conditions, it will inevitably cause immeasurable wear, scratches and other damage. In order to effectively detect the damage existing in the key parts of the tire, a tire surface damage detection method based on image processing was proposed. In this method, the image of tire side is captured by camera first. Then, the collected images are preprocessed by optimizing the multi-scale bilateral filtering algorithm to enhance the detailed information of the damaged area, and the optimization effect is obvious. Thirdly, the image segmentation based on clustering algorithm is carried out. Finally, the Harris corner detection method is used to capture the “salt and pepper” corner of the target region, and the segmsegmed binary image is screened and matched based on histogram correlation, and the target region is finally obtained. The experimental results show that the similarity detection is accurate, and the damage area can meet the requirements of accurate identification. Full article

Mural paintings, as the main components of painted cultural relics, have essential research value and historical significance. Due to their age, murals are easily damaged. Obtaining intact sketches is the first step in the conservation and restoration of murals. However, sketch extraction often suffers from problems such as loss of details, too thick lines, or noise interference. To overcome these problems, a mural sketch extraction method based on image enhancement and edge detection is proposed. The experiments utilize Contrast Limited Adaptive Histogram Equalization (CLAHE) and bilateral filtering to enhance the mural images. This can enhance the edge features while suppressing the noise generated by over-enhancement. Finally, we extract the refined sketch of the mural using the Laplacian Edge with fine noise remover (FNR). The experimental results show that this method is superior to other methods in terms of visual effect and related indexes, and it can extract the complex line regions of the mural. Full article

Radar data can provide additional depth information for monocular depth estimation. It provides a cost-effective solution and is robust in various weather conditions, particularly when compared with lidar. Given the sparse and limited vertical field of view of radar signals, existing methods employ either a vertical extension of radar points or the training of a preprocessing neural network to extend sparse radar points under lidar supervision. In this work, we present a novel radar expansion technique inspired by the joint bilateral filter, tailored for radar-guided monocular depth estimation. Our approach is motivated by the synergy of spatial and range kernels within the joint bilateral filter. Unlike traditional methods that assign a weighted average of nearby pixels to the current pixel, we expand sparse radar points by calculating a confidence score based on the values of spatial and range kernels. Additionally, we propose the use of a range-aware window size for radar expansion instead of a fixed window size in the image plane. Our proposed method effectively increases the number of radar points from an average of 39 points in a raw radar frame to an average of 100 K points. Notably, the expanded radar exhibits fewer intrinsic errors when compared with raw radar and previous methodologies. To validate our approach, we assess our proposed depth estimation model on the nuScenes dataset. Comparative evaluations with existing radar-guided depth estimation models demonstrate its state-of-the-art performance. Full article

Precise cloud and aerosol identification hold paramount importance for a thorough comprehension of atmospheric processes, enhancement of meteorological forecasts, and mitigation of climate change. This study devised an automatic denoising cloud–aerosol classification deep learning algorithm, successfully achieving cloud–aerosol identification in atmospheric vertical profiles utilizing CALIPSO L1 data. The algorithm primarily consists of two components: denoising and classification. The denoising task integrates an automatic denoising module that comprehensively assesses various methods, such as Gaussian filtering and bilateral filtering, automatically selecting the optimal denoising approach. The results indicated that bilateral filtering is more suitable for CALIPSO L1 data, yielding SNR, RMSE, and SSIM values of 4.229, 0.031, and 0.995, respectively. The classification task involves constructing the U-Net model, incorporating self-attention mechanisms, residual connections, and pyramid-pooling modules to enhance the model’s expressiveness and applicability. In comparison with various machine learning models, the U-Net model exhibited the best performance, with an accuracy of 0.95. Moreover, it demonstrated outstanding generalization capabilities, evaluated using the harmonic mean F1 value, which accounts for both precision and recall. It achieved F1 values of 0.90 and 0.97 for cloud and aerosol samples from the lidar profiles during the spring of 2019. The study endeavored to predict low-quality data in CALIPSO VFM using the U-Net model, revealing significant differences with a consistency of 0.23 for clouds and 0.28 for aerosols. Utilizing U-Net confidence and a 532 nm attenuated backscatter coefficient to validate medium- and low-quality predictions in two cases from 8 February 2019, the U-Net model was found to align more closely with the CALIPSO observational data and exhibited high confidence. Statistical comparisons of the predicted geographical distribution revealed specific patterns and regional characteristics in the distribution of clouds and aerosols, showcasing the U-Net model’s proficiency in identifying aerosols within cloud layers. Full article

Image fusion is an effective and efficient way to express the feature information of an infrared image and abundant detailed information of a visible image in a single fused image. However, obtaining a fused result with good visual effect, while preserving and inheriting those characteristic details, seems a challenging problem. In this paper, by combining a multi-level smoothing filter and regional weight analysis, a dual-band image fusion approach is proposed. Firstly, a series of dual-band image layers with different details are obtained using smoothing results. With different parameters in a bilateral filter, different smoothed results are achieved at different levels. Secondly, regional weight maps are generated for each image layer, and then we fuse the dual-band image layers with their corresponding regional weight map. Finally, by imposing proper weights, those fused image layers are synthetized. Through comparison with seven excellent fusion methods, both subjective and objective evaluations for the experimental results indicate that the proposed approach can produce the best fused image, which has the best visual effect with good contrast, and those small details are preserved and highlighted, too. In particular, for the image pairs with a size of 640 × 480, the algorithm could provide a good visual effect result within 2.86 s, and the result has almost the best objective metrics. Full article

Due to the difficulty in generating a 6-Degree-of-Freedom (6-DoF) object pose estimation dataset, and the existence of domain gaps between synthetic and real data, existing pose estimation methods face challenges in improving accuracy and generalization. This paper proposes a methodology that employs higher quality datasets and deep learning-based methods to reduce the problem of domain gaps between synthetic and real data and enhance the accuracy of pose estimation. The high-quality dataset is obtained from Blenderproc and it is innovatively processed using bilateral filtering to reduce the gap. A novel attention-based mask region-based convolutional neural network (R-CNN) is proposed to reduce the computation cost and improve the model detection accuracy. Meanwhile, an improved feature pyramidal network (iFPN) is achieved by adding a layer of bottom-up paths to extract the internalization of features of the underlying layer. Consequently, a novel convolutional block attention module–convolutional denoising autoencoder (CBAM–CDAE) network is proposed by presenting channel attention and spatial attention mechanisms to improve the ability of AE to extract images’ features. Finally, an accurate 6-DoF object pose is obtained through pose refinement. The proposed approach is compared to other models using the T-LESS and LineMOD datasets. Comparison results demonstrate the proposed approach outperforms the other estimation models. Full article

Motorcycle detection and collision warning are essential features in advanced driver assistance systems (ADAS) to ensure road safety, especially in emergency situations. However, detecting motorcycles from videos captured from a car is challenging due to the varying shapes and appearances of motorcycles. In this paper, we propose an integrated and innovative remote sensing and artificial intelligence (AI) methodology for motorcycle detection and distance estimation based on visual data from a single camera installed in the back of a vehicle. Firstly, MD-TinyYOLOv4 is used for detecting motorcycles, refining the neural network through SPP (spatial pyramid pooling) feature extraction, Mish activation function, data augmentation techniques, and optimized anchor boxes for training. The proposed algorithm outperforms eight existing YOLO versions, achieving a precision of 81% at a speed of 240 fps. Secondly, a refined disparity map of each motorcycle’s bounding box is estimated by training a Monodepth2 with a bilateral filter for distance estimation. The proposed fusion model (motorcycle’s detection and distance from vehicle) is evaluated with depth stereo camera measurements, and the results show that 89% of warning scenes are correctly detected, with an alarm notification time of 0.022 s for each image. Outcomes indicate that the proposed integrated methodology provides an effective solution for ADAS, with promising results for real-world applications, and can be suitable for running on mobility services or embedded computing boards instead of the super expensive and powerful systems used in some high-tech unmanned vehicles. Full article

Contrast enhancement techniques serve the purpose of diminishing image noise and increasing the contrast of relevant structures. In the context of medical images, where the differentiation between normal and abnormal tissues can be quite subtle, precise interpretation might become challenging when noise levels are relatively elevated. The Fast Local Laplacian Filter (FLLF) is proposed to deliver a more precise interpretation and present a clearer image to the observer; this is achieved through the reduction of noise levels. In this study, the FLLF strengthened images through its unique contrast enhancement capabilities while preserving important image details. It achieved this by adapting to the image’s characteristics and selectively enhancing areas with low contrast, thereby improving the overall visual quality. Additionally, the FLLF excels in edge preservation, ensuring that fine details are retained and that edges remain sharp. Several performance metrics were employed to assess the effectiveness of the proposed technique. These metrics included Peak Signal-to-Noise Ratio (PSNR), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Normalization Coefficient (NC), and Correlation Coefficient. The results indicated that the proposed technique achieved a PSNR of 40.12, an MSE of 8.6982, an RMSE of 2.9492, an NC of 1.0893, and a Correlation Coefficient of 0.9999. The analysis highlights the superior performance of the proposed method when contrast enhancement is applied, especially when compared to existing techniques. This approach results in high-quality images with minimal information loss, ultimately aiding medical experts in making more accurate diagnoses. Full article

Aiming at the problems of uneven light reflectivity on the spherical surface and high similarity between the stems/calyxes and scars that exist in the detection of surface defects in apples, this paper proposed a defect detection method based on image segmentation and stem/calyx recognition to realize the detection and recognition of surface defects in apples. Preliminary defect segmentation results were obtained by eliminating the interference of light reflection inhomogeneity through adaptive bilateral filtering-based single-scale Retinex (SSR) luminance correction and using adaptive gamma correction to enhance the Retinex reflective layer, and later segmenting the Retinex reflective layer by using a region-growing algorithm. The texture features of apple surface defects under different image processing methods were analyzed based on the gray level co-occurrence matrix, and a support vector machine was introduced for binary classification to differentiate between stems/calyxes and scars. Deploying the proposed defect detection method into the embedded device OpenMV4H7Plus, the accuracy of stem/calyx recognition reached 93.7%, and the accuracy of scar detection reached 94.2%. It has conclusively been shown that the proposed defect detection method can effectively detect apple surface defects in the presence of uneven light reflectivity and stem/calyx interference. Full article

Bridge inspection methods using unmanned vehicles have been attracting attention. In this study, we devised an efficient and reliable method for visually inspecting bridges using unmanned vehicles. For this purpose, we developed the BIRD U-Net algorithm, which is an evolution of the U-Net algorithm that utilizes images taken by unmanned vehicles. Unlike the U-Net algorithm, however, this algorithm identifies the optimal function by setting the epoch to 120 and uses the Adam optimization algorithm. In addition, a bilateral filter was applied to highlight the damaged areas of the bridge, and a different color was used for each of the five types of abnormalities detected, such as cracks. Next, we trained and tested 135,696 images of exterior bridge damage, including concrete delamination, water leakage, and exposed rebar. Through the analysis, we confirmed an analysis method that yields an average inspection reproduction rate of more than 95%. In addition, we compared and analyzed the inspection reproduction rate of the method with that of BIRD U-Net after using the same method and images for training as the existing U-Net and ResNet algorithms for validation. In addition, the algorithm developed in this study is expected to yield objective results through automatic damage analysis. It can be applied to regular inspections that involve unmanned mobile vehicles in the field of bridge maintenance, thereby reducing the associated time and cost. Full article

In digital image processing, filtering noise is an important step for reconstructing a high-quality image for further processing such as object segmentation, object detection, and object recognition. Various image-denoising approaches, including median, Gaussian, and bilateral filters, are available in the literature. Since convolutional neural networks (CNN) are able to directly learn complex patterns and features from data, they have become a popular choice for image-denoising tasks. As a result of their ability to learn and adapt to various denoising scenarios, CNNs are powerful tools for image denoising. Some deep learning techniques such as CNN incorporate denoising strategies directly into the CNN model layers. A primary limitation of these methods is their necessity to resize images to a consistent size. This resizing can result in a loss of vital image details, which might compromise CNN’s effectiveness. Because of this issue, we utilize a traditional denoising method as a preliminary step for noise reduction before applying CNN. To our knowledge, a comparative performance study of CNN using traditional and embedded denoising against a baseline approach (without denoising) is yet to be performed. To analyze the impact of denoising on the CNN performance, in this paper, firstly, we filter the noise from the images using traditional means of denoising method before their use in the CNN model. Secondly, we embed a denoising layer in the CNN model. To validate the performance of image denoising, we performed extensive experiments for both traffic sign and object recognition datasets. To decide whether denoising will be adopted and to decide on the type of filter to be used, we also present an approach exploiting the peak-signal-to-noise-ratio (PSNRs) distribution of images. Both CNN accuracy and PSNRs distribution are used to evaluate the effectiveness of the denoising approaches. As expected, the results vary with the type of filter, impact, and dataset used in both traditional and embedded denoising approaches. However, traditional denoising shows better accuracy, while embedded denoising shows lower computational time for most of the cases. Overall, this comparative study gives insights into whether denoising will be adopted in various CNN-based image analyses, including autonomous driving, animal detection, and facial recognition. Full article

During the service process of high-rise buildings, hollowing defects may be produced in the decorative layer, which not only affect the appearance, but also create a safety hazard of wall covering and shattered plaster peeling. Numerous studies have shown that hollowing can be detected using infrared thermal imagery under normal conditions. However, it is difficult to detect the edge and calculate the area of the hollowing on an exterior facade accurately because of the low contrast and fuzzy boundaries of the obtained infrared thermal images. To address these problems, a method for extracting the contours of building facade hollowing defects using polarization thermal images based on an improved Canny algorithm has been proposed in this paper. Firstly, the principle of thermal polarization imaging was introduced for hollowing detection. Secondly, considering the shortcomings of the Canny edge detection algorithm and the features of polarization thermal images, an improved Canny edge detection algorithm is proposed, including adaptive bilateral filtering to improve noise reduction ability while ensuring defect edges are not virtualized, Laplacian sharpening and histogram equalization to achieve contour sharpening and contrast enhancement, and eight-direction gradient templates for calculating image gradients, which make interpolation with non-maximum suppression more accurate, and the Tsallis entropy threshold segmentation algorithm based on the OTSU algorithm verification makes the image contour information more complete and accurate. Finally, a long-wave infrared polarization thermal imaging experimental platform was established and validation experiments were conducted. The experimental results demonstrate that the distinct, smooth, and precise location edges of the hollowing polarization infrared thermal images can be obtained, and the average error of the detected hollowing area is about 10% using the algorithm proposed in this paper. Full article