节点文献
含倾斜文字的图像垃圾邮件过滤技术研究
Research on Image with Skew Document Spam Filtering Technology
【摘要】 垃圾邮件制造者采用图像来隐藏垃圾信息,把文字隐藏在图像中并带一定的倾斜。在对含有倾斜文字的图像邮件深入分析后,针对图片中含有倾斜文字的垃圾邮件,文章给出了一种在进行OCR识别之前对图像进行hough变换,并结合支持向量机识别垃圾邮件的算法,实验结果表明,该方法具有比较满意的过滤效果。
【Abstract】 Spammers use images to hide the spam,the text is hidden in images,and with a certain degree of skew.This paper give in-depth analysis of images containing skew document.A method that prior to OCR recognition images hough transform,combined with support vector machine algorithm for spam recognition has been presented.Experimental results show that this method has satisfactory filtering effect.
【关键词】 图像垃圾邮件;
Hough变换;
光学字符识别;
支持向量机;
【Key words】 image spam; hough trasformation; optical character recognition; support vector machine;
【Key words】 image spam; hough trasformation; optical character recognition; support vector machine;
【基金】 成都理工大学青年基金资助
- 【文献出处】 计算机与数字工程 ,Computer & Digital Engineering , 编辑部邮箱 ,2010年05期
- 【分类号】TP393.098
- 【被引频次】4
- 【下载频次】75