Characters Extraction from Strings on a Document Image Using Handwritten Marks on Touch Screen

We argued validity of Tablet PCs in the fields of E-learning, and this paper discussed a new character extraction system from strings on the document image using handwritten marks on touch screen. As the first step of this study, we proposed the method to identify handwritten notes/marks to associate strings on the documents with handwritten marks. In this paper, experiments using actual scanned documents were also conducted. As an experimental result, 90% of notes extraction rate was obtained and 60% of marks were appropriately extracted. We confirmed that the proposed method would be a useful user interface for the character extraction from the documents. Keywords-Tablet Device, E-learning, Electronic Documents