ReP-ETD: A Repetitive Preprocessing technique for Embedded Text Detection from images in spam emails

Email service proves to be a convenient and powerful communication tool. As internet continues to grow, the type of information available to user has shifted from text only to multimedia enriched. Embedded text in multimedia content is one of the prevalent means for delivering messages to content viewers. With the increasing importance of emails and the incursions of internet marketers, spam has become a major problem and has given rise to unwanted mails. Spammers are continuously adopting new techniques to evade detection. Image spam is one such technique where in embedded text within images carries the main information of the spam message instead of text based spam. Currently, image spam is evaluated to be roughly 50% of all spam traffic and is still on the rise, thus a serious research issue. Filtering mails is one of the popular approaches used to block spam mails. This work proposes new model ReP-ETD (Repetitive Pre-processing technique for Embedded Text Detection) for efficiently and accurately detecting spam in email images. The performance of the proposed ReP-ETD model has been evaluated across the identified parameters and compared with other existing models. The simulation results demonstrate the effectiveness of the proposed model.