Method for identifying image of junk e-mail based on high-order autocorrelation characteristic

Abstract

The invention relates to an image processing technique and a correlative technique in the field of mode identification and discloses a method and a device for automatically identifying the image of a junk e-mail. In the invention, the automatic classification of the image of a normal e-mail and the image of a junk e-mail is realized by utilizing the high-order autocorrelation characteristic of the image and combining a support vector machine. The method comprises the following steps: firstly, an edge detection operator is utilized to extract the binaryzation edge of the image; secondly, 25 high-order autocorrelation characteristic extraction templates of 3*3 are designed according to a high-order autocorrelation function and are utilized to obtain the 25-dimensional high-order autocorrelation characteristic of the image; thirdly, normalization processing is carried out to the obtained 25-dimensional high-order autocorrelation characteristic, and all the characteristic values are positioned between 0 and 1; and finally, automatic identification is carried out to the image by utilizing a classifier of the support vector machine. In the invention, because the high-order autocorrelation characteristic has the advantages of translation and rotational invariance, the device has better anti-interference performance.

Description

technical field [0001] The invention relates to image processing and pattern recognition technology, in particular to a method for feature extraction and recognition of spam images. Background technique [0002] After more than 30 years of development, e-mail has become an indispensable part of the Internet, and more and more people use e-mail as a tool for their communication and communication. However, the proliferation of spam has made this convenient means of communication a huge challenge. At present, my country has become one of the hardest hit areas of spam. According to a survey report released by the Anti-Spam Center of the Internet Society of China in January 2008, more than half of the emails received by Chinese users are spam. The problem of spam has attracted the attention of all walks of life, and a large number of spam detection algorithms have emerged, such as IP blacklists, behavior-based filtering methods, rule-based spam detection algorithms, text conten...

AI Technical Summary

Problems solved by technology

Although this method has a faster recognition speed, it can only identify fewer image spam

Images