Imaging techniques for secure photography using dual camera

Abstract

With the increase in consumer generated image content using handheld devices such as mobile cameras and ease of availability and operability of image editing software, there is an acute need for methods for secure photography. These methods should help us address issues such as ownership management, authentication, and forgery detection. This paper aims to propose an algorithm for secure photography by embedding photographer's face (or related features) as an invisible watermark into the images captured by a smartphone equipped with dual cameras. The proposed system embeds an image taken by front camera while photographing a scene from the back camera. The proposed algorithms are first profiled in Matlab to make them ready for implementation on the Android phone. The trade-off between data size, quality of the owner's image, and the quality of the scene images is used to design an optimal operating point for the system. The results shown in this paper show that the proposed method holds promising grounds for addressing the issues related to ownership management of mobile images.