On Digital Watermarking Algorithms For Vector Geospatial Data

Geospatial vector data is the core in the applications of geographic information systems. However, digital format geospatial data in network environment has made it easy to copy and illegally distribute geospatial data. It is not enough to only depend on traditional security measures to stop piracy of geospatial data; therefore a more reliable and effective technology is needed to explore.In recent years, digital watermarking which is an advanced technique in information security field has appeared. It embeds watermark information in the digital products, and makes it a part of the vector data that cannot be separated, thereby to judge who is the copyright owner and who is the pirate. Digital watermarking is playing a more and more important role in the geospatial vector data security and copyright protection.The characteristics of digital watermarking for geospatial vector data are analyzed in this paper. Combining with GIS theories and geospatial vector data features, a series of geospatial vector data watermarking algorithms are studied based on normalization method, the mapping function, the Discrete Fourier Transform, mathematical statistics and other mathematical theories. The main research results of this paper are summarized as follows:(1) Based on the organization and structural characteristics of geospatial data, the features of the digital watermarking technology for geospatial vector data are analyzed, and the watermark attacks on geospatial vector data and the impact on robustness of watermarking algorithm are summarized.(2) Two different watermarking algorithms for geospatial vector data in spatial domain are proposed considering characteristics that watermark attacks on geospatial vector data. By means of embedding watermark information into min-max normalized spatial data, a blind watermarking algorithm based on data normalized for geospatial vector is proposed; on the basis of introducing QR code technology and an improved algorithm based on the traditional LSB, a blind watermarking algorithm for geospatial vector data is addressed.(3) Aiming at the key problems of watermarking algorithm in DFT domain for geospatial vector data, three blind watermarking algorithms for vector geospatial data based DFT are proposed. Through study and analysis of errors of data caused by watermarking algorithm based on DFT, a data error control scheme for watermarking algorithm based on DFT is designed and implemented and a blind watermarking algorithm for vector geospatial data under controllable errors based on DFT is proposed. As the poor robustness to vertex attack of conventional watermarking algorithm based on DFT, a watermarking algorithm based on the feature point and DFT is proposed; the transform domain algorithm cannot be directly applied to the vector point data. In this algorithm vector point data is divided into cells by regular grid, and a virtual line feature is composed of all points of every cell, and a DFT domain watermarking algorithm for point data based on grid partitioning is proposed.(4) Advantages and disadvantages of the spatial domain and DFT domain watermarking algorithm are analyzed. Based on multi-technology integration strategy, through effectively combing multiple watermarking method making up for each other’s deficiencies and resist more watermarking attacks, a multiple blind watermarking algorithm based on spatial-frequency domain for vector map data is proposed. By this means, the robustness of watermarking algorithm is improved.(5) According to projection transformation and coordinate transformation attacks that are unique in geospatial vector data, a robust to projection attacks blind watermarking algorithm for vector data is proposed with full consideration of features in the projection transformation and coordinate transformation of geospatial vector data. In the watermark embedding procedure, the WGS84 geographic coordinates system which is in common use in an intermediate coordinate system. A watermark is embedded into the WGS84 coordinate space data. When the watermark detecting the data must be converted from the original vector space coordinates to WGS84 coordinate system, the watermark can be detected and no original data is needed. The algorithm is robust against projection transformation and coordinate transformation attacks and also against other attacks such as cropping, rotation and translation.