Information embedding in real object image using temporally brightness-modulated light

Abstract

We propose a new technology that can be used to invisibly embed information into the images of real objects that are captured with a video camera. This technique uses illumination that invisibly contains certain information. Because the illumination on a real object contains information, an image of the object taken with a video camera also contains information although it cannot be seen in the captured image. This information can be extracted by image processing. It uses temporally luminance modulated patterns as invisible information. The amplitude of the modulation is too small to perceive. The frequency of modulation is the same as the frame frequency of the projector that is used as a lighting device. The frame images over a certain period are added up after the sign of the even- or odd-numbered frames is changed. Changes in brightness by modulation in each frame are accumulated over the frames. However, the object and background image are removed because the even and odd frames are opposite in sign. As a result, the patterns become visible. We conducted experiments and the results from these revealed that invisible patterns could be read out. Moreover, we evaluated the invisibility of the embedded patterns and confirmed that conditions existed where both the invisibility and readability of the patterns were simultaneously satisfied.