Haptic Rendering for Direct Hand Interaction
Abstract
Computer graphics has satisfactorily shown its ability to describe and visually render complex and diverse interactive digital scenarios for multiple applications. Thanks to virtual reality, users can enter these computer-generated worlds, enabling an extension to the range of applicability. One of the most relevant accomplishments of this technology lies in the possibility to spatially interact with objects in the environment. The user is part of the scene, hence he/she can interact with the scene and feel immersed. Together with virtual reality, other spatial computing technologies share the challenge of providing a natural interface, with which the user can interact intuitively and which, in turn, also provides reliable haptic feedback. Motivated by this challenge, this work proposes methods that augment the interaction and the immersion of the user in virtual reality and other immersive technologies, by allowing direct touch interaction, thanks to which the user can realistically interact with the virtual environment while perceiving corresponding haptic feedback. First, we formulate a physics-based soft hand simulation model that follows the user’s hand and allows realistic dexterous manipulation of virtual objects. Later, we propose a novel optimization-based tactile rendering approach for wearable devices that displays haptic feedback that matches the stimulus computed in the virtual interaction, based on an accurate simulation of contact between a virtual finger and the virtual environment. Finally, we study the problem of haptic rendering for underactuated haptic devices. We carry out a theoretical analysis to study the passivity of the system considering the user’s intent, and we leverage our findings to design a novel rendering method. Overall, the presented computational models set up the basis for a haptically rich bidirectional interface with the virtual environment, which empowers and enriches the user experience.
Description
Tesis Doctoral leída en la Universidad Rey Juan Carlos de Madrid en 2021. Director/es: Miguel Ángel Otaduy
Collections
- Tesis Doctorales [1495]