A Geographical Databases Watermarking Method Based on Quantization Modulation with Variable Steps

Geographical databases, as one kind of the most important infrastructure of geographical information system, are great treasure of data owners. How to protect the copyright of geographical data effectively using digital watermarking is a critical issue. In this paper, we propose a robust, shape-preserving and blind watermarking method. We compute mean feature distance for each polygon and choose h most significant bits of mean feature distance as the robust identifier of one polygon. All polygons are partitioned into several groups based on their identifiers. Quantization modulation technique with variable steps is exploited to hide watermark into all polygons whose areas are slightly modified to derive watermarked geographical database. To ensure the security of the proposed algorithm, the polygon-group assignment, the watermark bit to be embedded in each polygon and every variable step are all algorithmically determined under the control of a private key known only to the owner of the data. Experimental results show that the proposed watermarking method has good performance and is resilient to translation, rotation, simplification, noise addition, vertices interpolation, cropping, tuple alteration and insertion attacks. Moreover, within the usage range of the geographical data, the robustness of the proposed method is improved with the increase of the watermark strength.