Hiding Visual Patterns in Halftone Images

Abstract

To get higher contrast between the hidden visual pattern and the background of decoded image is one of the most important objectives in the visual cryptography research for halftone images. Pei and Guo proposed a noise-balanced error di®usion method (NBED) for hiding visual patterns in halftone images and had a nice experimental result. But in their method, since the noise-balanced value in°uences the quality of images apparently, it is tuned and decided by the user. In this paper, we propose a method named "Automatic Noise-Balanced Error Di®usion" (ANBED) which determines the noise-balanced values automatically. Our method generates two images: the traditional error di®usion image (TED) and a hidden visual pattern image (HV P). When the TED is superimposed by the HV P, the hidden visual pattern can be seen by human eyes. Simulation result shows that our automatic method is very close to Pei and Guo0s method with the best tuned noise-balanced value.