Intuitive Input Methods for Interactive Segmentation on Mobile Touch-Based Devices

Existing interactive image segmentation approaches mainly focus on the algorithms rather than the input methods. The literature regarding the input methods is mainly applied to desktop PCs. However, there is a transition to mobile touch based devices. In this paper we describe our approach to identify a set of intuitive input methods for interactive segmentation on mobile devices. Preliminary results of two user studies are presented. A description of our planned research is given, divided into initial and refinement input methods and the exploitation of the input for segmentation algorithms.