An Accurate Authentication Algorithm Based on Point Constraint Block for Vector Geographic Data

Fragile watermarking technique has been viewed as an effective method to achieve content authentication, which not only detect any modifications that occurred, but also locate the modified areas. Based on fragile watermarking technology, an accurate authentication scheme based on point constraint block is proposed, which detects the malicious attacks with high accuracy while ensuring to locate exactly the tampered area for vector geographic data. Our innovative strategy is based on point constraint block and uses the spatial relationships between the elements of data points. In the authentication information embedding process, vector geographical data is divided into blocks according to the method of point constraint block, and the spatial positional relationship between the data points of each block is reordered by the "Zig-Zag" pattern, so as to establish an organized positional relationship between data points. Then, for each point, the fragile watermark information is generated by its adjacent point, and is embedded into the current point. In the process of content authentication, the extracted watermark information is compared with the generated watermark information, and the comparative result is used to judge whether the data have been updated. The proposed algorithm is furthermore compared with the method that is based on a uniform block, and the experimental results show that the proposed authentication algorithm can accomplish accurate authentication when the data is updated, and it has the ability to achieve the accurate authentication of deleted elements. Meanwhile, when data modifications are detected, it can locate and mark the modification positions.