Bankura Unnayani Institute of Engineering

This paper presents a reproducible experimental and statistical methodology for evaluating the thermo-mechanical response of marble and granite subjected to extreme temperature conditions. The method is designed to support rock mechanics applications in fire-affected structures, geothermal systems, and underground engineering. Two varieties of Makrana marble and two types of granite were systematically heated from 35 °C to 600 °C using a controlled thermal protocol. Changes in density and uniaxial compressive strength were measured following standardized testing procedures to quantify thermal damage and mechanical degradation. A statistical analysis framework, including regression modeling and Pearson’s correlation analysis, was implemented to characterize the relationships between temperature and rock properties. The proposed methodology enables consistent assessment of thermal sensitivity in crystalline rocks and can be readily adapted for other lithologies and high-temperature engineering scenarios. This work provides a practical reference for researchers seeking standardized experimental and analytical approaches to evaluate rock performance under elevated temperatures.

Floods are a recurring natural disaster in India, causing significant damage to life and infrastructure. This paper presents a machine learning-based approach to predict flood levels in already flooded areas, using image data from the ISRO Bhuvan website, weather data based on geographic coordinates, and elevation data. To get better accuracy of the proposed model, a convolutional neural network (CNN) is combined with the Long Short-Term Memory (LSTM) networks. A combination of CNN and LSTM is used to process the image and give it to LSTM along with numerical weather data. The experimental results show that the proposed model achieves an accuracy of 98

Industry 4.0 represents a transformative shift in the field of manufacturing and industrial processes that are associated with the interconnection of cyber-physical systems, the Internet of Things (IoT), and sophisticated data analysis. In this regard, predictive maintenance has become a vital approach to improve the efficiency of the operation process, minimize downtimes, and increase the lifetime of industrial resources. With the large volumes of data created each second by the IoT-based sensors, predictive maintenance uses data mining algorithms to find the trends and anomalies that can predict the possibility of equipment malfunction before it happens. Compared to the traditional reactive or scheduled maintenance, which is inactive and reactive, this proactive approach allows for smarter decisions and allocation of resources more optimally. An end-to-end data mining system for predictive maintenance in an Industry 4.0 IoT environment includes data collection over a variety of sensor networks, data pre-processing to guarantee the quality of the data, scalable storage systems, sophisticated machine learning algorithms to make accurate predictions, and visualization tools to facilitate maintenance scheduling and operational control. With the help of these elements, industries will be able to move to more robust and intelligent maintenance systems based on the objectives of Industry 4.0.

Medical imaging is pivotal in modern healthcare, offering a visual window into the human body’s intricate structures and functions. However, the scarcity of diverse and representative medical images significantly limits research progress. Today, deep learning algorithms are increasingly incorporated into the medical imaging domain to automate the diagnostic process. The success of these algorithms relies heavily on vast and diverse datasets. Insufficient data hampers the training and validation of these algorithms, resulting in suboptimal performance, biased results, and reduced generalizability. This paper introduces a pioneering approach that employs a Convolutional Autoencoder (CA) to synthetically generate medical images. The synthetic images produced are then added to the existing database to create an augmented dataset. This augmented dataset is subsequently used for classification with a convolutional neural network. Experiments were conducted on publicly available datasets—the chest CT-scan dataset and the IQ-OTH/NCCD lung cancer dataset. The synthetic image generation capability of the CA was compared with traditional augmentation methods such as flipping, rotating, shearing, shifting, zooming, and sub-sampling. Results showed that the CA-based augmented dataset achieved an accuracy of 91 %, compared to 83 % with the traditional augmentation-based dataset.

Digital image authentication is a critical process in ensuring the integrity and ownership of digital images, especially in the age of widespread digital manipulation and unauthorized use. In this manuscript, an image watermarking scheme has been proposed by combining the Lifting Wavelet Transform (LWT) and the Least Significant Bit (LSB) for image authentication. The goal of the proposed scheme is to achieve robustness against various attacks while maintaining the high-quality visual appearance of the watermarked image. Experimental evaluations are conducted to assess the performance of the proposed hybrid watermarking scheme. Various metrics, including Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Normalized Correlation (NC), are employed to measure the visual quality, robustness, and authentication accuracy of the watermarked image. The results demonstrate the effectiveness of the proposed scheme in achieving high-quality visual appearance, robustness against attacks, and accurate image authentication.