Detecting Erase Strokes from Online Handwritten Notes Using Support Vector Classification

We have implemented a student note-sharing system, AirTransNote, that facilitates collaborative and interactive learning in conventional classrooms. With the AirTransNote system, a teacher can immediately share student notes with the class using a projection screen to enhance group learning. However, students tend to hesitate to share their notes, particularly when the notes contain embarrassing mistakes. Nevertheless, teachers want to focus on real mistakes students make while learning. We introduce an erase stroke detecting method for the student note-sharing system to reduce students’ discomfort regarding sharing mistakes, as well as to assist the teacher in finding mistakes. We collected and manually labeled free-style handwritten student notes. Based on the labeled notes, we extracted features for the erase symbols and deleted strokes. We have tested support vector machine techniques for classifying erase symbols and deleted strokes from typical handwritten notes. c ⃝ 2015 The Authors. Published by Elsevier B.V. Peer-review under responsibility of KES International.