Evaluation of a Pointing Interface for a Large Screen Based on Regression Model with Image Features

Abstract

Gesture input interfaces for mobile devices with a touch screen have become widespread. Although gesture interfaces in common use are limited to the small screens of these mobile devices, pointing interfaces for large screens using handheld devices or attachments are common. However, these devices increase the user's cognitive load because they are unable to cancel the effect of spatial cognition. This paper presents a pointing interface for a large screen. This interface is based on learning a user's pointing motion and providing for intuitive pointing by canceling the effect of spatial cognition. Multiple linear regression analysis is used for learning and estimation. This method was implemented in a prototype system with a large projected screen, and experiments with the prototype system confirmed the validity of the proposed method.