为解决基于数字水印的无线多媒体消息版权管理系统对提取后水印标识的自动识别问题，本文在充分考虑多媒体消息在传播中可能遭受攻击的基础上，提出一种基于Gabor小波特征的标识确认方案。该方案利用这类小波函数确定的滤波器适合局部分析和多方向多尺度分析的特点，提取与水印版权标识结构信息相关的统计量，形成特征集向量，通过特征集的距离比较，在小尺寸水印质量退化情况下，实现了对水印标识的识别。分析和实验表明，该方案能够满足无线多媒体消息版权管理的需求，并且在文中分析的情况下，设备的自动识别精度可以达到95%以上，较好地支持了对无线多媒体消息的版权管理。

To solve the problem of automatically identifying the copyright labels in case of adopting watermarking in the copyright management of wireless multimedia messages (MM), this paper, by considering the possible attacks on the watermarks, proposes an identification method based on Gabor wavelet characteristics. Since the Gabor wavelet filtering has good performance in local, multi-directional and multi-scale analysis, to exploit the advantages, the proposed scheme uses the filtering to extract the set of characteristics of copyright labels and identifies it by computing and comparing the distances between the set and the stored ones. The analysis and experiments show that the method can meet the requirements of MM’s copyright management, and, with the accuracy of identification be more than 95% under the conditions presented, well support the copyright management of multimedia messages.