摘要

The copyright protection of two-dimensional (2D) vector map has attracted a lot of research focus due to the increasing security issues raised in recent years. One promising direction seeking the optimal tradeoff between adding watermarks and maintaining minimal distortion is the so-called lossless watermarking, i.e., after watermark extraction the 2D vector maps are fully lossless. This paper presents a novel lossless watermarking scheme for 2D vector maps based on a novel recursive embedding algorithm. In our algorithm, feature points of individual polylines are first grouped into united, upon which highly correlated unites are selected as cover data to carry out a recursive modification of its mean vertex coordinates. Such operation not only ensures lossless compression, but also enables higher payload capacity and, to a certain degree, the perception invisibility before and after the watermark extraction. We have conduced experiments on several real-world 2D vector map applications to show the effectiveness, efficiency of the proposed algorithm.