A scaling parameter optimization of watermarking using autostereograms as random dot images

For a digital watermarking, some autostereograms are used as random dot images. In an autostereogram image space every dot is randomly distributed, therefore the extraction efficiency is good and the distortion rate is low. In this paper, we develop a method to choose an optimized scaling parameter at which the distortion is very low and the extraction efficiency is very high for watermarking of some arbitrary general images.