The Study Of Adaptive Blinding Watermarking Algorithm In DC/AC Coefficient

With rapid development of digital technology and daily popularization of Internet, the exchange of digital product are becoming easier, at the same time, the copyright protection question is more serious. How to use the convenience of Internet, and protect the intellectual property rights, brings to people more and more attention. The digital watermark technology is proposed to solve this difficult problem, and becomes a hotspot rapidly. After summarizing some typical algorithms, this thesis proposes two kind of watermarks algorithms, which insert the watermark in gradation pictures and text documents, concrete work as follows:Firstly. It introduces the merit and the general steps of the DCT watermark algorithm. Then proposes one kind of adaptive watermark algorithm based on the quantification remainder, and discusses key questions of the algorithm in detail: chooses the insert position, confirms the feasibility of inserting the watermark into the DC coefficients by an experiment; chooses the embedment strength, analysis some factors and the concrete experiment, has produced the method of determination the embedment strength when the carrier is gradation pictures. This algorithm improves the blemish which is not blind, at the same time it keeps from the block effect which brought in by changing too much of DC coefficients. It combines DC and AC coefficient to embed. It embeds 1-bit watermark into 4 coefficients by setting remainder, and connects the 4 coefficients to strengthen its robustness. By discussing HVS and attacks may meet, chooses the adaptive strength of embedding, It has a better invisibility and robustness. The experimental result has proved it can survive JPEG compression, and several kinds of image processing attacks.Secondly. Summarizes characteristic of two values pictures (text) and several conclusions of watermark algorithm in text documents. Proposes one kind of text watermark algorithm, and analysis several important parameters of some steps to study their influence for algorithm In the capacity, invisibility and robustness aspects discusses this algorithm. At last gives some advice to use it.