SVM-Driven Approach for Source Camera Identification

This research study presents a Support Vector Machine (SVM)-driven approach for source camera identification, focusing on extracting Sensor Pattern Noise (SPN) as a distinctive feature. The study investigates the effectiveness of SVM in accurately identifying the source camera by leveraging SPN and other intrinsic image characteristics. Comparative analysis with advanced models, including Random Forests, Gradient Boosting Machines, and Convolutional Neural Networks (CNNs), highlights the strengths and limitations of SVM. While CNNs demonstrate higher accuracy, SVM provides an efficient balance between accuracy and computational resources. The proposed SVM model is evaluated on publicly available datasets and achieves robust performance, demonstrating its relevance for forensic applications, security, and image authenticity verification. Further improvements are discussed, including strategies for enhancing model accuracy and handling complex classification scenarios.