The clutch: two-handed mobile multi-touch 3D object translation and manipulation

Nowadays, handheld devices such as smartphones provide users with multi-touch input screens. Displaying interactive and touch-enabled 3D environments in such handheld devices has become popular in different applications like games or virtual reality. Technologies such as Web3D and WebGL have made the creation and display of 3D environments in mobile devices easier than ever. However, object manipulation techniques are not as well developed. For example, moving an object within the 3D environment or other similar object-specific manipulations are neither intuitive nor easy to perform. Current manipulation techniques like Gizmo that are successful in systems that use mouse and keyboard are not designed for and do not work well for multi-touch handheld devices. In this paper, we present a novel technique to perform object manipulation in 6DOF in multi-touch screens. Our performance evaluations show that our technique compared to existing techniques such as Gizmo improves task completion time by 63% while increasing task precision by 52%.

I. Introduction

Today, mobile 3D multimedia is widely used in a variety of applications such as mobile gaming, medical simulations, multimedia enhanced learning and working, virtual museum, tourist guides, and other interactive simulations. Many studies have reported that simplicity and usability are important factors for the adoption and acceptance of mobile 3D applications. However, designing simple and effective techniques for 3D interaction on touch-based mobile devices is a challenging task. While many research works have studied this issue, some interaction aspects need further research. In this paper, we introduce a novel method for 3D object translation and manipulation that is especially useful when two-handed mobile touch is used as the only input modality, which is very common today with smartphones or tablets. As an example, consider the case when a player wants to move a gun inside a mobile 3D game. In this situation, the player needs a specific interaction means to move the gun within the 3D space. While this is trivial on a gaming console or a device that has mouse and keyboard, it is very challenging in a touch display input device. We will show in the next chapter that most existing methods proposed for 3D object translation are not suitable for multi-touch input with two fingers on handheld devices. The main reason for this is that the user's input must take place on a small 2D plane (the display), yet moving and rotating virtual 3D objects requires user input with 6 degrees of freedom (6DOF).