基于灰度—梯度共生矩阵的图像型垃圾邮件识别方法  

Image spam identification method based on gray-gradient co-occurrence matrix

作　　者：冯兵[1,2] 李芝棠[1,2,3] 花广路 

机构地区：[1]华中科技大学计算机科学与技术学院,湖北武汉430074 [2]下一代互联网接入系统国家工程实验室,湖北武汉430074 [3]华中科技大学网络与计算中心,湖北武汉430074

出　　处：《通信学报》2013年第S2期1-4,共4页Journal on Communications

摘　　要：为了逃避基于文本的垃圾邮件系统的检测,越来越多的垃圾邮件制造者将文本信息嵌入到图像中。为了有效地检测出图像型垃圾邮件,提出了一种基于灰度—梯度共生矩阵(GGCM,gray-gradient co-occurrence matrix)的图像型垃圾邮件识别方法。先通过灰度—梯度共生矩阵提取图像的特征信息,然后运用最小二乘支持向量机(LS-SVM,least squares support vector machines)进行分类。实验表明,该方法具有较高的分类精度和较好的实时性。In order to avoid the detection of the spam system based on text, more and more spammers have embedded text information into the image. An image spam identification method based on gray-gradient co-occurrence matrix(GGCM) was proposed to detect image spam effectively. The feature of image was extracted through GGCM firstly, and then LS-SVM was used to do classification. The test results show that this method has higher classification accuracy and better real-time performance.

关 键 词：图像型垃圾邮件 灰度—梯度共生矩阵 最小二乘支持向量机 纹理特征 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]