Research On Multiple Dynamic Digital Watermarking Algorithms For Vector Geographic Data

Vector geographic data are very important basic and strategic national information resource,which play a significant role in the national economy and infrastructure construction.The safety of vector geographic data is closed with state and national security.With the advent of the informational era,it becomes much easier to acquire,transmit and distribute vector geographic data.The security problems become increasingly prominent.Furthermore,the protection demands of multiple users and multi-copyright for vector geographic data which are caused by multi-level transmission is increasingly urgent.It is currently urged to focus on how to effectively protect the multi-copyright of vector geography data and track multiple users,especially track the dynamic using of multiple users.Multiple watermarking technology is that embedding multiple watermarks into the cover data,and it can be used to protect multi-copyright of multiple users and track the copying flow of data in the progress of multi-level transmission.Therefore,multiple watermarking technology can effectively solve the above information security problems of vector geographic data.This article,which analyzes the characteristics of vector geographic data and multiple watermarking technology,focuses on the research of multiple dynamic digital watermarking algorithms for vector geographic data.The main research work and achievements are as follows:(1)Based on the data features of vector geographic data and multiple watermarking technology,technical characteristics of multiple digital watermarking algorithms are concluded,multiple watermarking attacking methods and applications of multiple digital watermarking technology for vector geographic data have been also proposed.And then the two different watermarks embedding models,"one to one" and "many to one" model,are put forward according to the mapping relationships between watermark bits and vector geographic data.They are the basis of designing the multiple watermarking algorithms for vector geographic data.(2)Based on the "one to one" mapping model and focusing on improving the robustness against the cropping attacks,the multiple digital watermarking algorithms for vector geographic data based on dividing domains are designed.Multiple watermarks are embedded into the sub-blocks of dividing logic domains,and the multiple watermarking is designed based on mapping and dividing for vector geographic data.The algorithm has good performance on robustness against cropping attacks.The other multiple digital watermarking is designed based on correlation detection and discriminant analysis,the multiple watermarks are randomly embedding into the coordinates and detected according to correlation detection and discriminant analysis,which has good robustness and is easily implemented.(3)Based on the "many to one" mapping model and focusing on removing the conflicts between the watermark bits in the same coordinate,the multiple digital watermarking algorithms for vector geographic data based on superposition are designed.The multiple digital watermarking algorithm is designed according to adding embedding rule and correlation detection.The algorithm is suited for the data with small data size.The other multiple digital watermarking is designed based on multiple quantizations.The multiple watermark bits are embedded into different quantification areas.The algorithm has the followed advantages:good robustness,large watermark capacity,and accuracy controlability.(4)Focus on the "dynamic" of multiple dynamic digital watermarking for vector geographic data,the self-detection algorithms of multiple digital watermarking for vector geographic data are designed.Based on analyzing the statistical property of the extracted watermark bits set corresponding to the single watermark bit index,the single watermark bit detection model is built,and then the self-detection algorithm based on statistical property is designed.The watermarks can be detected without the original data and watermarks and has good robustness.According to the statistical property of the extracted watermarks and focus on the above problem,a self-detection statistic count is built and the robust self-detection algorithm is designed.The algorithm has good robustness,is suit for the data with small data size and blind detection.The self-detection algorithms provide the important conditions for the watermarks embedding dynamically.