International Journal of Computing and Related Technologies

Enhanced Visual Cryptography Based on Arnold's Cat Map

Hajir Alauldeen Al-Bayati, Dalal N Hamod, Lahieb M Jawad — Mon, 20 Jan 2025 15:38:46 +0000

The field of color visual cryptography has experienced significant growth, enabling the secure transmission of color images over the Internet. This innovative methodology partition images into shares, mirroring the principles of Naor and Shamir's approach but tailored specifically for color images. Color visual cryptography (VC) operate by divide color secret images into shares, treat each color pixel individually through the RGB and CMY channels. This study introduces an advanced CVC technique that integrate a chaotic encryption system, significantly enhance security measures. The proposed approach involves extract the channels component and apply a chaotic map to each channel with distinct masks, resulting in the generation of six shares (two for each color channel), subsequently XORing them with a randomly generated key matrix. This process creates new matrices, further enhance security. On the recipient's end, the secret color image is retrieved while maintaining high quality, as evaluated through metrics such as Peak Signal-to-Noise ratio PSNR, Mean Square Error MSE, Correlation Coefficients CC, Number of Pixels Changing Rate NPCR, and Unified Average Change Intensity UACI. Comparative analysis against existing methods demonstrates the superior efficient of this methodology in securely concealing color image secrets. Where the values of each metric were CC =1.0, MSE= 0.0 and PSNR= infinite. While, the best values of NPCR and UACI were 99.65%, 85.62% respectively these for the original channel and associated shares.

Image Denoising Using Multi-Model Fusion Technique

Kamal Khan, Muhammad Anwar, Saifullah Khan — Mon, 20 Jan 2025 15:44:26 +0000

Image denoising is a fundamental challenge in the field of image processing, with the primary goal of recovering high-quality images from noisy counterparts. This paper investigates the effectiveness of multimodal fusion techniques for denoising images. The study utilizes the Waterloo Exploration Database, a comprehensive collection of 4,744 pristine natural images, selecting 500 images for experimentation. Gaussian noise was artificially introduced to simulate realistic noise conditions, creating the noisy input for the denoising process. Multiple modalities—grayscale, edge, and depth images—were extracted from the noisy images to capture different aspects of the visual content. These modalities were aligned and combined using early fusion techniques, producing a single cohesive representation. A Convolutional Neural Network (CNN) was then trained using this fused data for image denoising. The evaluation focused on key metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and Mean Squared Error (MSE). The results indicate that multimodal fusion significantly improves denoising performance, as evidenced by increased PSNR and reduced MSE, suggesting its potential to enhance image restoration methods.

The Green Drive: A Comparative Analysis of Carbon Emissions of Traditional Fuel-Based Vehicles

Mon, 20 Jan 2025 15:53:19 +0000

This research exposition delves into the ecological ramifications of gasoline-fueled automobiles vis-à-vis their gasoline-driven counterparts. This investigative endeavor employed a comparative methodology, mindful of the escalating apprehensions surrounding climatic alterations and the imperative for sustainable conveyance. The compilation encompasses a plethora of vehicles utilizing carbon monoxide, accentuating authentic emissions data, life cycle examinations, and manufacturers. Participants were meticulously chosen from an array of designs and models to furnish a paradigmatic emissions framework emblematic of vehicular advancement. Procuring emissions data from esteemed enterprises, ecological collectives, and peer-reviewed sources constituted the inaugural phase of the data accumulation process. Despite assiduous endeavors to ascertain precision, limitations of recognition encompass discrepancies in reportage methodologies and biases inherent in the data proffered by corporations, underscoring the import of gauging outcomes in academic milieus. Model Accuracy - Linear Regression 348.24, Ridge Regression 345.45, Lasso Regression 337.21, KNN Regression 314, SVR Regression 333.12, Random Forest 268.4 Actual Value is 28 - Random Forest predication is near to more accurate than other applied models.

Edge Computing: Applications & Challenges: A Short Review

Dua Noor, Imran Khan, Shahbaz Qamar, Shahzad Ayaz — Mon, 20 Jan 2025 16:02:46 +0000

Edge computing is distributed computing system in which Edge devices are used to process data locally, at the point of generation. Compared to alternative computer architectures, it offers higher latency as data processing and analysis are completed on the premises. Input-output ports, memory, storage, CPUs, and other components are integrated into edge computing tools. On these devices, data processing and analysis programs are installed at the location of data creation. A distributed computing paradigm known as "edge computing" improves reaction times and conserves bandwidth by bringing processing and data storage closer to the point of demand. Low latency, real-time processing, and the capacity to handle data at the edge of the network are important features. Edge computing is useful for many applications, including Internet of Things (IoT), virtual and augmented reality, smart cities, and driverless cars. Local data processing lowers latency, making it appropriate for applications where timely decision-making is essential. This study is to focus applications and challenges of edge computing along with tools which imparts key part of the edge computing. These tools and techniques can lead a better way to improve edge computing challenges. Hence this study recommends few research direction for research with regard latest development. This paper is review in the field edge computing.

Machine Learning-Based Prediction of Wind Power Using MATLAB

Omar Ratib Khazaleh, Firas Ratib Alawneh, Ali Ratib Alkhuzaie — Mon, 20 Jan 2025 16:09:58 +0000

The increasing pressure to increase the use of renewable energy in the face of climate change has pushed the sustainable energy agenda forward with wind power at the forefront due to the fact that it has a large greenhouse gas emission reduction curve. However, due to the variation in wind speed, its penetration challenges undermine the predictability of power systems, which is essential for integrating volatile renewable energy sources into grids. In this paper, the focus is on the application of machine learning (ML) algorithms in improving the availability of reliable wind power forecasts as applied to wind power management and integration into power systems. In this context, this paper presents the post hoc analysis through different neural network configurations such as feedforward neural networks (FNNs), nonlinear autoregressive networks with external inputs (NARX), back-propagation neural networks, and radial base function (RBF) networks. These models were trained and validated for their ability to predict wind energy using data obtained from the Fujairah wind turbines in Jordan, which were obtained through a SCADA system from November 2013 to August 2024. The accuracy evaluation of the constructed models shows that the proposed FNN model, due to its ease and efficiency in determining the number of layers and neurons, showed the smallest RMSE and MAE compared to other models, thus confirming the FNN model as the most accurate and reliable model for wind energy prediction. On the other hand, the RBF network, which this study observed its ability to resolve nonlinear data, provided less positive results, indicating that the algorithm will require further optimization and further combination with the specific nonlinearity of wind energy data. From this study, we can highlight the importance of machine learning techniques when it comes to wind energy prediction, which is very important for wind energy management and exploitation. The knowledge derived from this research enhances the theoretical understanding of previous work in this area, but also provides compelling recommendations for energy suppliers seeking to enhance the effectiveness of their business operations and integrate more sustainable supply methods.