Comics reading: An automatic script generation

Abstract

With the advent of portable devices, reading comic ebooks is a popular activity. However, a simple scan of a comic page is not well adapted for portable device screens and a panel to panel reading without animations and adapted transitions is quite uncomfortable and not suitable. Moreover, applying manually transitions between each panel to script a complete comic book is a tricky task and seems impossible for a complete collection of comics. We present a model able to automatically script comics reading by using panel lines of force. Our results demonstrate that this model proposes a coherent solution for 87.2% of panels in an interactive time.