Watermarking 2D-vector data for geographical information systems

This paper deals with the issue of watermarking 2D-vector data which are used in Geographical Information Systems (GIS). The watermark is embedded in the tolerance range of the coordinates, where one bit of the watermarking information is represented by one PN-sequence, whose elements consist of the two values +tolerance and - tolerance. To robustly embed one bit of the watermarking information the length of the PN-sequence has to be much greater than the square maximum coordinate value leading to non-acceptable sequence lengths due to high coordinate values. To achieve a PN-sequence length that is suitable to the size of the data domain we do not consider the whole coordinate value but only those decimal digit positions of the coordinate value, where changes are significant but not violating the tolerance requirements. Due to this restriction on a smaller range of values, overflow and underflow has to be considered during the embedding process. Within the retrieval process we first extract this fraction of the coordinate value before correlating it with the PN- sequence. The proposed method is robust against attackers changing the coordinates within the tolerance range.

[1]  Yacov Yacobi,et al.  Improved Boneh-Shaw Content Fingerprinting , 2001, CT-RSA.

[2]  John G. Proakis,et al.  Probability, random variables and stochastic processes , 1985, IEEE Trans. Acoust. Speech Signal Process..