The emerging field of haptics, which enables the sensing and manipulation of virtual environments through touch, is recognized as an important element in the area of human computer interaction. Many of its applications, including medical training, rehabilitation, security, entertainment, etc. involve the transmission, storage, protection, and analysis of haptic information. However, the multi-dimensional and time-varying nature of haptic data along with the fact that high sampling frequencies are required when generating haptic signals, motivate the research of haptic data reduction. This thesis discusses two important and related topics: data reduction and protection in multimodal visual-haptic virtual environments.;The protection of haptic information through digital watermarking is also investigated in this thesis due to its conceptual similarity to haptic data reduction and compression. A foundational study is conducted to examine the role of multisensory feedback in the perception of a watermark embedded in a haptic-enabled 3D virtual surface. In particular, the analysis aims to answer numerous important questions in this brand-new research area, including: Do watermarks inspected using multimodal feedback (vision + haptic) result in very different detection thresholds from those detected using a single sensory modality; or more importantly, whether multimodal visual-haptic feedback improves the perception of watermarks embedded in 3D meshes.;The focus of the first topic is on the design, evaluation, and analysis of haptic data reduction techniques in order to: (1) Improve packet transmission in telehaptic systems, and (2) reduce the dimensionality of the inherently large haptic datasets. In the former case, robust perception-based data reduction methods in 3- and 6-degree-of-freedom telehaptic systems are derived. The proposed techniques are evaluated in different experimental settings, including a haptic-enabled telesurgery simulation, and demonstrate a significant reduction in haptic data traffic while maintaining a high-quality telehaptic experience. In the latter case, data reduction in haptic information analysis problems is investigated. Specifically, data reduction methods are presented within the context of haptic data mining and knowledge discovery to help facilitate the analysis of the inherently high-dimensional haptic datasets. Experiments using datasets of haptic-based handwritten signatures demonstrate the accuracy of the techniques in reducing the high dimensionality of haptic feature spaces.
展开▼
