Melanoma, the deadliest form of skin cancer, is on the rise. The goal of this study is to present a deep learning system implementation for the detection of melanoma lesions on a server equipped with a graphics processing unit (GPU). When applied by a dermatologist, the recommended method might aid in the early detection of this kind of skin cancer. Evidence shows that deep learning may be used in a variety of settings to successfully extract patterns from data such as signals and images. This research presents a convolution neural network–based strategy for identifying early-stage melanoma skin cancer. Images are input into a deep learning model known as a convolutional neural network (CNN) that has already been pre-trained. The CNN classifier, which is trained with large amounts of data, can discriminate between malignant and nonmalignant melanoma. The method's success in the lab bodes well for its potential to aid dermatologists in the early detection of melanoma. However, the experimental results show that the proposed technique excels beyond the state-of-the-art methods in terms of diagnostic accuracy.

Medical images are a specific type of image that can be used to diagnose disease in patients. Critical uses for medical images can be found in many different areas of medicine and healthcare technology. Generally, the medical images produced by these imaging methods have low contrast. As a result, such types of images need immediate and fast enhancement. This paper introduced a novel image enhancement methodology based on the Laplacian filter, contrast limited adaptive histogram equalization, and an adjustment algorithm. Two image datasets were used to test the proposed method: The DRIVE dataset, forty images from the COVID-19 Radiography Database, endometrioma-11, normal-brain-MRI-6, and simple-breast-cyst-2. In addition, we used the robust MATLAB package to evaluate our proposed algorithm’s efficacy. The results are compared quantitatively, and their efficacy is assessed using four metrics: Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE), Contrast to Noise Ratio (CNR), and Entropy (Ent). The experiments show that the proposed method yields improved images of higher quality than those obtained from state-of-the-art techniques regarding MSE, CNR, PSNR, and Ent metrics.

Skin cancer is among the most widely distributed, deadliest cancers around the globe, and early diagnosis becomes vital to enhance patient survival. Deep learning has demonstrated high potential for automatic skin lesion classification. However, existing Convolutional Neural Networks (CNNs) are still unsatisfactory in terms of dataset reliance, subjection to orientations, and the inability to model long-range global context. To solve this problem, we propose a hybrid model named the VGG19-RSPDA-ViT with the fine-grained local feature captured by VGG19 and the global context provided by Vision Transformers (ViT). The proposed RSPDA enforces rotation invariance and enriches the feature space, which further strengthens generalization on a small training set. To the best of our knowledge, this is the first work to systematically combine feature-map-level rotation/shift augmentation with a CNN-Transformer hybrid model for dermoscopic skin cancer detection. Performance was validated on three benchmark datasets: the Melanoma Skin Cancer Dataset of 10,000 Images (MSK10000) for binary classification and the Human Against Machine with 10,000 training images (HAM10000) and Hospital Pedro Hispano (PH2) for multi-class classification. Our model achieved accuracies of 97.9%, 97.1%, and 98.67% on MSK10000, HAM10000, and PH2 datasets, respectively, with consistently high macro-averaged precision, recall, specificity, and F1 scores across both datasets. VGG19-RSPDA-ViT outperformed existing state-of-the-art methods with superior generalization capabilities. These results demonstrate that our proposed model is effective for skin lesion classification and has significant potential for clinical application as an automated diagnostic tool in dermatology.

Image classification and feature extraction have been studied extensively and used efficiently in several applications. This paper suggests a novel method by combining three main methods for texture feature extraction. The proposed method is based on Local Ternary Pattern (LTP), Texture Feature Coding Method (TFCM), and Gray Level Cooccurrence Matrix (GLCM). We have entitled our method as GCLTP which is stand for Gray Coding Local Ternary Pattern. The combination of LTP, TFCM, and GLCM is assigned a unique value used to extract the features of an image. GCLTP is tested using images are taken from the Brodatz database. A set of 22 features were extracted from images. GCLTP is experimentally accomplished a high accuracy in classification by using the most known classifiers.

As the world continues to battle the devastating effects of the COVID-19 pandemic, it has become increasingly crucial to screen patients for contamination accurately and effectively. One of the primary screening methods is chest radiography, utilizing radiological imaging to detect the presence of the virus in the lungs. This study presents a cutting-edge solution to classify COVID-19 infections in chest X-ray images by utilizing the Gray-Level Co-occurrence Matrix (GLCM) and machine learning algorithms. The proposed method analyzes each X-ray image using the GLCM to extract 22 statistical texture features and then trains two machine learning classifiers - K-Nearest Neighbor and Support Vector Machine - on these features. The method was tested on the COVID-19 Radiography Database and was compared to a state-of-the-art method, delivering highly efficient results with impressive sensitivity, accuracy, precision, F1-score, specificity, and Matthew's correlation coefficient. The proposed approach offers a promising new way to classify COVID-19 infections in chest X-ray images and has the potential to play a crucial role in the ongoing fight against the pandemic.

The Internet of Things (IoT) is a cutting-edge innovation that facilitates the cost-effective development of smart system architectures. Although current regulations necessitate installing an analog fire alarm system, such a system lacks the intelligence to instantly notify the appropriate parties in a timely fashion. In addition, since people are not always present, an analog fire alarm will not be able to prevent immediate danger or damage in the event of a fire. Therefore, the incident must be reported as soon as possible to the appropriate party in order to lessen the impact of a fire. In this study, we suggest a smart fire-alarm system made of a fire sensor and a sound sensor that can both detect fire and noise as well as the status of the analog fire alarm system to ascertain whether the analog fire alarm system is operational. We first tested our proposed smart fire alarm system to determine its effectiveness before putting it into use. From there, we ran experiments to determine how well it worked. The outcomes show that the system is trustworthy in a range of scenarios.