CamDetector: Detecting and Positioning Hidden Cameras with Raspberry Pi 4 via EM Radiation

The illicit use of spy cameras poses a significant threat to scenarios where privacy is paramount, such as in hotel rooms and home rentals. Owing to advancements in miniaturization and camouflage techniques, these covert cameras have become increasingly difficult for victims to detect in their surroundings. In this paper, we introduce a proof-of-concept system called CamDetector, designed to identify spy cameras based on their electromagnetic (EM) radiations. Our initial step involves characterizing the EM radiation emitted by cameras. We then uncover the shared pattern stemming from the signal coupling on the CSI cable, which can be used to detect hidden cameras. We use USRP B200 mini and Raspberry Pi 4 to build the prototype and proceed to assess the performance on 6 commercially available small cameras. The experimental results demonstrate the effectiveness of CamDetectorwith an average accuracy of 90.08% and a recall of 88.21% at a distance of 0.4m. Moreover, our evaluation based on user involvement confirmed that CamDetector can achieve 85% accuracy in an
$18 m^{2}$
room, with an average time of 16.9 minutes.