A simple sheet bending recognition for augmenting a two-dimensional marker-based response analyzer

Abstract

Conventional response analyzers use various student devices such as clickers and mobile devices, which require the additional burdens of charging and managing. Fiducial marker-based response analyzers have been proposed to relieve these burdens. The IDs and rotations of the fiducial markers are used to identify students and their respective answers. The marker-based approach is straightforward, but transferrable data are limited to the ID and the rotations. To enhance the transferrable data, we introduce a method of extracting curves of the marker sheet. Since a paper sheet is flexible, students can control the shape of the sheet intuitively. We have implemented our method by modifying a conventional fiducial marker recognizer and confirmed its effectiveness.