Image Transformation for IoT-based Anomaly Detection

In the rapidly evolving environment of the Internet of Things (IoT), anomaly detection plays an important role in ensuring the security and reliability of connected systems. Traditional approaches often struggle to capture subtle patterns and anomalies in complex IoT data. This paper introduces a new perspective by utilizing image transformation techniques, specifically the Gramian Angular Field (GAF), Markov Transition Field (MTF), and Recurrence Plot (RP), coupled with deep learning models for better anomaly detection. This study presents a comparative study to evaluate the performance of these image transformation methods and assess their effectiveness in detecting anomalies in IoT data. The comparative results show that RP with CNN has the superior performance compared to other image transformation techniques.