Intelligent Mouse-based Object Group Selection Inproceedings

Hoda Dehmeshki, Wolfgang Stuerzlinger


Modern graphical user interfaces support direct manipulation of objects and object groups. Current object group selection techniques such as lasso and rectangle selection can be time-consuming and error-prone. This paper presents a new approach to group selection that exploits the way human perception naturally groups objects, also known as Gestalt grouping. Based on known results from perception research, we present a novel method to group objects via models of the Gestalt principles of proximity and (curvi-)linearity. Then, we introduce several new mouse-based selection techniques that exploit these Gestalt groups. The results of a user study show that our new technique outperforms lasso and rectangle selection for object groups with an implicit structure, such as (curvi-)linear arrangements or clusters.

Date of publication: Aug - 2008
Get PDF Get Video Get Citation