Shape-Preserving Algorithm forWatermarking 2-D Vector Map Data

A novel statistic based algorithm for watermarking 2D vector maps is proposed. Unlike other multimedia data types, 2D vector map data has many special features, which make the selection of the cover data become a crucial issue directly influencing the performance of the watermarking scheme. In our proposed scheme, the selected cover data is a distance sequence extracted from the original map based on a feature point detecting procession. For a certain original map, this distance sequence can be regarded as a random variable with a stable distribution and the watermark is embedded by changing its distribution. Watermark extraction is based on statistic detection. Experiment results show that the proposed method has better shape-preserving performance, which is of great importance for vector maps. Moreover, this scheme is proved robust to certain attacks such as map interpolation, simplification, data reordering and some geometric transformations like rotation, translation, and their combination