Location Spoofing Detection for Social Network Service using Camera Fingerprint
Qianru Zhao, Ning Zheng, Qing Bao · 2020
Check-in on location-based social network services have become a popular trend recently. Meanwhile, people may deceive their location to seek illegal violence. Making the data reliability of location-based social networks negatively affected. Hence, it is important to identify the authenticity of location. However, methods in literature rely heavily on user-generated content and check-in data, and these methods may be invalidated due to missing prior information or lacking rich training samples. To address this challenge, this paper proposes an innovative method for detecting the location spoofing based on the source of the image for specific scenarios. Specifically, we first extract the camera sensor fingerprint based on the images posted by an inquiry user via the well-designed denoising filter. Second, the authenticity of the location is tested by comparing the consistency of the residual noise from newly-posted images with location check-in and the unique camera sensor fingerprint from an inquiry user. Finally, we conducted wide experiments on the image database, which empirically showed that the method we proposed is effective and simple, that is, it need no prior information, but only sample images from the social networks.