Halftone-Based Data Hiding Method Using Orientation Modulation Strategy

In smart cognitive cities, hiding data in printed matter is desirable in forensic printer applications, e.g. source tracking, authentication, and copyright control. In addition, the popularization of smart mobile devices totally changes the manner in which we communicate with the world around us. For example, by scanning the Coca-Cola can, people can easily obtain abundant on-line information. It is because the website information is hidden on packing paper of the can. In this paper, we focus on the class of clustered-dot halftones because it is preferable to electrophotographic printers due to stable dot formation. Therefore, given an arbitrary grayscale image, we design an embeddable cell selection scheme for the corresponding dithered halftone, and perform orientation modulation (OM) to the dot-cluster in the selected cells to embed information. To improve the quality of an output data-embedded halftone without damaging the embedded data, we present a modified direct binary search (DBS) framework by introducing a new search strategy. Hence, the proposed method integrates OM-based embedding with DBS quality optimization. Simulation results indicate that the proposed method surpasses other approach in both image quality and data capacity.