Fragile Word Document Watermarking Algorithm Based On Critical-Character

With the rapid development of computer network information technology, more and more text documents need to be transmitted on the Internet, and meanwhile network information transparency and operability brought security issues to the text content. Text watermarking technology as a branch of the digital watermarking technology, domestic and foreign scholar’s research is still small.Text document is different from other digital multimedia like image, video, audio, which contains less redundant information, so the progress of research is relatively small. In this paper, we use the special attributes of character, combined with the contents of the text document to realize text watermarking algorithm. The thesis of this paper is to realize the text watermarking algorithm’s tamper detection and recovery function.The watermark information which is generated based on the content in existing document fragile watermarking algorithm, each character is acted as the same method for processing. In fact, In fact, each character in the document to maintain the importance of the authenticity of the document is different, so a fragile Word document watermarking scheme based on critical character double verification is proposed. In this method, all characters in the Word document are divided into the critical character and non-critical ones. The8bits watermark data generated by each character are embedded in the special properties of the same character for the purpose of resisting general tampering. In order to improve the ability of tamper detection performance for the critical characters, the8bits watermark data generated by the critical character act as the second watermark data, which are embedded in the special properties of the other non-critical character, so it can realize the double verification of the critical character. In addition, we also discuss the impact that watermark data embedded in the special properties of the character on the size of Word document by statistical experiments. All those experimental results show that the watermark data embedded in the special properties can improve the security of watermarked Word document, and the dual watermark data which are embedded in critical character can improve the tamper detection performance greatly.In order to recovery the tampered character, a Word document recovery algorithm based on key character’s Unicode is proposed. The algorithm extracts the character’s Unicode then change into binary value as the recovery watermark information. Then we classify characters into3categories based on the significance of character to Word document. They are the most critical character and mid-critical character and times of key characters and non-critical character. The recovery watermark information of the three types key characters are embedded in a non-critical character and some of the key characters. On the basis of tampering detection algorithm, we recover he tampered character by extracting its recovery information. The experimental results show that the algorithm has high invisibility and can do best recovery on the tampered character.