Research And Realization On Copyright Protection Of The Engieering Drawing Based On PKI And Digital Watermarking

With the quick development of Internet, the engineering drawing is extensively applied. At the same time, it results in the problems about the copyright protection of the engineering drawing. The digital watermarking technique is an available tool to solve these problems. Presently, the research on digital watermarking focuses on the gray image, color image, audio and video digital watermarking, but there are little research on two-dimensional vector graphics and binary image digital watermarking. Moreover, the copyright protection of digital production like the engineering drawing not only needs the digital watermarking technique, but also requires the technique of Public Key Infrastructure (PKI). This thesis combines these two techniques, and design an architecture of copyright protection of the engineering drawing based on PKI and digital watermarking. The copyright protection system based on this architecture can resolve a series of problems about the copyright registration, trade, copyright authentication, copyright dispute and so on within the copyright protection of the engineering drawing.This thesis aims at the characteristics of two-dimensional vector graphics and proposes the two-dimensional vector graphics watermarking algorithm based on discrete wavelet transform. The algorithm withdraws the vertices coordinates which are generated into some sequences from the two-dimensional vector graphics. Then the sequences are transformed into different spatial and frequency coefficients through discrete wavelet transform. The algorithm embeds the watermarks into the two-dimensional vector graphics by modifying the wavelet magnitude coefficients. And lots of experiments have shown the good imperceptibility and robustness under the attacks.This thesis aims at the characteristics of binary image and proposes that the binary image watermarking can be realized by using the "zero-watermarking" technique. "Zero-watermarking" technique makes use of the important character of the image to construct watermarking information, and wouldn't change any information of the original image. This thesis proposes the binary image zero-watermarking algorithm based on discrete wavelet transform and invariant moments by using the present zero-watermarking algorithms for reference. Compared with the present zero-watermarking algorithms, the ultimate dissimilarity in this algorithm is to construct a meaningful zero-watermarking image. A lot of experiments have shown the good robustness and low false alarm rate.