2014 | OriginalPaper | Buchkapitel
Image Classification Method Using Support Vector Machine for Privacy Incident Response System
verfasst von : Jong-Bae Kim
Erschienen in: Future Information Technology
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Nowadays, Internet becomes more and more popular in our modern life due to the rapid development of information and communication technology. Internet is widely used in a vast of social activities, such as electronic commerce, Internet banking, which are based on a cyber space. To organize these social activities as well as to possess conveniences on the Internet, an electrical media expressed by personal identification certificates (resident cards, driving licenses, passports, etc.) is employed. As a result, it is necessary to generate an exposure of personal information on the web. Therefore, this paper proposes an efficient personal image classification method using support vector machine (SVM) for privacy incident response system(PIRS). In our case, PIRS detects the image included personal information. To extract the optimal features of the image included personal information, the proposed method selects common features from the training set. And then a personal image in the Web is detected using optimal image features and an SVM classification. The experimental results of the proposed method, the classification success rate of personal image is about 82%, and the miss-classification rate is about 13%. The cause of the miss-classification is because the images were digitized by various illuminations. However, the personal images for personal identification showed high classification rate.