Research Of Digital Image Watermarking And Authentication Technique Based On EZW

With the development of computer networks,multimedia technique and information science, copying of digital production, such as image,sounds,text or video etc, become so easier that the problem of copyright protection of multimedia products stares us in the face, and now we need to solve how to insert authentication information into digital production. Digital watermark technology is an effective solution to offer copyright protection and authentication, which has been produced in the past few years, more and more international scientists have been exerting their efforts in this field. Digital watermark technology has been one of the hottest subjects in the current academic world. Digital watermark has many advantages that differ from image encryption technique, but they complement each other a lot.This paper aims to implement image copyright protection and authentication. Firstly, this paper introduces the basic model frame of image digital watermark. This paper analyzes current processing technology for image digital watermark ,which includes creation,embedment,extraction,detection and authentication. We summarize current research status quo in this field. Then we narrate signal wavelet processing and image processing in wavelet domain. After we explain Embedded Zerotree Wavelet (EZW) coding, we take Human Visual System (HVS) and Watermark Assault Model into account and put forward a new watermark algorithm for static grew image, through which we can implement watermark embedment and detection. At last we propose a half frangible watermark algorithm based on the theory of digital signature, which can authenticate the integrity of image.We have mainly completed the following work in this paper:We have browsed many intestine and international literature concerning image watermark. We point out current research status quo in image watermark, we also classify different watermark techniques according to their basic character. We discuss the basic frame of image watermark and main research direction in the future.We analyze the theory and progress of wavelet transform and dispose image watermark through Discrete Wavelet Transform (DWT),at the same time many scholars have been doing this through Discrete Cosine Transform(DCT).DWT has three main advantages over DCT guaranteeing that image watermark won't be eliminated if the original image is compressed; wavelet has the peculiarity of visual multiresolution which is consistent with HVS,so we can choose appropriate position and intensity to embed watermark information; we can embed watermark in image compressing domain.We embark on thorough research for EZW technique and advance a watermark algorithm based on important coefficients of wavelet tree. Low frequency coefficients are always obviously eliminated in former watermark algorithm based on wavelet transform. In this paper we suggest that we should choose important coefficients with intermediate frequency or low frequency firstly, then embed watermark information in each kind of coefficients thus watermark will have good robustness and image quality.We reference the theory of digital signature to bring forward a half fragilewatermark algorithm based on image content which can be used for authenticating the integrity of image. Experiment has indicated that the algorithm can gain its ends and can be used for authentication of news photos,medical digital photos ,network business and transmission application.