随着多媒体网络通讯技术的飞速发展，数字信息的安全维护问题日益突出.目前，采用数字水印技术 (即脆弱性数字水印和半脆弱性数字水印 )进行数字图象的真实性、完整性认证已成为信息认证领域的研究热点.为使国内广大科技人员能够较全面地了解数字水印图象认证技术的发展现状，给出了用于图象认证的数字水印系统的基本框架、性能要求及常见的攻击方法，介绍了现有的各种算法，分析和总结了各自的优缺点，并提出了下一步的研究方向.

The growth of networked multimedia systems has made digital data acquisition, exchange and transmission a simple task, but the ease of copying and editing also facilitates unauthorized use, misappropriation and misrepresentation, which makes it necessary to authenticate the multimedia data. At present, it has been a hotspot to authenticate digital images by the use of watermarking. According to the objectives of authentication, an image authentication system can be classified into complete verification and content verification. Watermarking for complete verification (known as fragile watermarking) considers image data as untouchable messages such that the data for authentication have to be exactly the same as the original. Content verification is a characteristic of multimedia data authentication. Watermarking for content verification (known as semi fragile watermarking) considers "information preserving" image manipulations, such as compression and format conversion, as acceptable. This paper presents the general framework of digital watermarking for image authentication, discusses the fundamental demands and common attacks, introduces various existing algorithms and analyzes their advantages and disadvantages. The paper also introduces the current state of the techniques and proposes several research topics at next stage.