Freehand Grasping: An Analysis of Grasping for Docking Tasks in Virtual Reality

Natural and intuitive interaction in VR as grasping virtual objects, is still a significant challenge and while recent studies have begun to explore interactions that aim to seamlessly create virtual environments that mimic reality as closely as possible, the dexterous versatility of the human grasp poses significant challenges for usable and intuitive interactions. At present the design considerations for creating natural grasping based interactions in VR are usually drawn from the body of historical knowledge presented for real object grasping. While this may be suitable for some applications, recent work has shown that users in VR grasp virtual objects differently than they would grasp real objects. Therefore, these interaction assumptions may not be directly applicable in furthering the natural interface for users of VR, presenting an absence of knowledge on how users intuitively grasp virtual objects. To begin to address this, we present two experiments where participants (N=39) grasped 16 virtual objects categorised by shape in a mixed docking task exploring rotation, placement and target location. We report on a Wizard of Oz methodology and extract grasp types, grasp category and grasp dimension. We further provide insights into virtual object categorisation for assessing interaction patterns and how these could be used for developing natural and intuitive grasp models by parameterizing grasp types found in these experiments. Our results are of value to be taken forward into a framework of recommendations for grasping interactions and thus begin to bridge the gap in understanding natural grasping patters for VR object interactions