Design and implementation of image SPAM filtering and analysis system
Wang Xing-wei · Dalian Ligong Daxue xuebao · 2011
The image SPAM is filtered by starting with the basic features of image and selecting some simple image attributes as the filtering features.To enhance the filter performance,the edge gray point rate feature is proposed which reflects the text information in the image.Then,K-means is used to solve the problem of dividing intervals and the rough feature interval(RFI) and purified feature interval(PFI) are got.Finally,a fast filtering mechanism based on pure feature intervals is advanced.And for rough feature intervals,a mechanism based on support vector machine(SVM) is proposed,by which the accuracy rate is 98.396 6%.All the training features used in the filtering model can be extracted in only one scan of the image,so the filtering mechanism can meet the needs of system efficiency.