Robust Video Surveillance Imagery Reconstruction Via Electromagnetic Leakage
Electronic cameras may help realize real-time guards, yet it can also be exploited by eavesdroppers to capture private information. This study presents a novel side-channel attack that utilizes electromagnetic radiation (EMR) signal by electronic cameras to infer the visual content displayed on surveillance screens. We intend to map between the video data stream and the EMR signal to recover the original video image. As an important technical contribution of this work, we build a video reconstruction algorithm that extracts the original monitoring images from the leaked EMR signal output. The results show that we can restore surveillance screen content with extremely high fidelity, even in challenging real-world environments.