In print-type steganography and watermarking, visible calibration patterns are arranged around content in which invisible data are embedded, to provide plural feature points for normalization of a scanned image. However, it is clear that conventional visible calibration patterns interfere with the page layout and artwork of original contents. Additionally, visible calibration patterns are not suitable for security services. In this paper, we propose an arrangement of and a detection method for an invisible calibration pattern based on human visual perception. We embed the proposed calibration pattern in an original image by adding a high-frequency component to blue intensity in limited regions. The proposed method protects the page layout and artwork because it is difficult for observers to detect the calibration pattern embedded into an image with a normal background.
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition
- Computer Science Applications
- Computer Graphics and Computer-Aided Design