Title :
Application of Scale Invariant Feature Transform to Image Spam Filter
Author :
Chen, Junwei ; Zhang, Lichun ; Lu, Yue
Author_Institution :
Dept. of Comput. Sci. & Technol., East China Normal Univ., Shanghai
Abstract :
Inspired by the keyword-based text filter, this paper proposes an image filter which detects the spam image by matching with user-specified image content. In this way, detecting image spam e-mail is converted into image matching process. Stable local feature detection and representation is a fundamental component of image matching algorithms. SIFT has been proven to be the most robust local invariant feature descriptor. In this process, SIFT algorithm is applied. The images are extracted with SIFT features, which are used to carry out the image matching work. Our experiments demonstrate that SIFT has a good performance in spam image recognition.
Keywords :
feature extraction; image matching; image representation; transforms; unsolicited e-mail; feature extraction; image matching process; image recognition; image spam filter; keyword-based text filter; local feature detection; local invariant feature descriptor; scale invariant feature transform; spam e-mail; user-specified image content; Advertising; Character recognition; Computer vision; Electronic mail; Feature extraction; Image matching; Image recognition; Matched filters; Robustness; Unsolicited electronic mail;
Conference_Titel :
Future Generation Communication and Networking Symposia, 2008. FGCNS '08. Second International Conference on
Conference_Location :
Sanya
Print_ISBN :
978-1-4244-3430-5
Electronic_ISBN :
978-0-7695-3546-3
DOI :
10.1109/FGCNS.2008.24