压缩域互联网信息监测过滤仪关键技术研究  

Research on Key Techniques of Compressed Domain Networking Information Surveillance and Filtering Instrument

作　　者：沈兰荪[1] 王素玉[1] 

机构地区：[1]北京工业大学信号与信息处理研究室,北京100022

出　　处：《现代科学仪器》2006年第5期44-48,共5页Modern Scientific Instruments

基　　金：国家自然科学基金(No.60402036;No.60431020;No.60301003);北京市自然科学基金(No.4042008;4062006);教育部博士点基金(No.20040005015)

摘　　要：针对互联网信息监测管理的需求,我们提出基于压缩域的网络信息监测过滤仪的概念,并从压缩域纹理信息的分类、轮廓提取、字符定位、肤色分割、人脸检测等几个方面对其所涉及的关键技术进行了广泛深入的研究,提出了一系列高效的压缩域图像处理算法,为监测过滤仪的进一步开发设计奠定了基础。On meeting the requirement of networking information surveillance and management, the concept of compressed domain networking information surveillance and filtering instrument was proposed. Furthermore, the key techniques of it were investigated widely and deeply from the aspects of texture information classification, contour extraction, text location, skin color segmentation and face detection etc. A series of effective processing algorithms of compressed domain image are proposed, providing the foundation for further designation of the instrument.

关 键 词：DCT压缩域 小波变换域 纹理信息分类 字符定位 肤色分割 人脸检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]