Abstract
Two-dimensional vector maps are vulnerable to issues related to copyright owner's protection, which have an impact on the vector-data usage in different applications. Watermarking schemes are required in order to prevent this type of attack rotation, uniform scaling and translation (RST), invariance property, embedding distortion control and a high level of capacity in their concealment of a digital vector map's information. This process typically involves transforming an original map and embedding the watermark. In this paper, a new approach to copyright protection for vector maps is proposed using a linear cellular automata (LCA) algorithm. The approach taken is as follows: (1) an original map, which is LCA-transformed, is obtained; (2) the watermark bit insertion process is conducted on the coefficient of the transformation result frequency into the LSB pattern; and (3) the inverse LCA transformation of the map is employed to obtain the watermarked map. Subsequent analysis has found that this approach provides desirable levels of fidelity, invisibility, capacity and computational complexity. It is also resistant to geometric and signal-operations attacks and provides multi-frequency domains for digital watermarking.