Techniques for Sketch-Based User Interface: Review and Research

This paper summarizes the state of the art of the techniques on online sketchy shape recognition and semantic understanding for sketch-based user interface. Firstly, several existing methods for online sketchy shape recognition are classified into stroke-based, primitive-based, feature-based and composite-based, and anatomized respectively. Secondly, some typical ways to sketch understanding, including semantics acquisition, semantic interpretation and semantic application, are expounded. Finally, some important open issues and relevant solutions of sketch-based user interface are synthetically explored from the point of view of ambiguity of geometry, adaptation of use and independency of application respectively.