Research On Flexible Physical Tactile Rendering In Virtual Reality

Haptic deformation rendering as an important part of virtual reality, plays an important role in improving the sense of reality, immersion and interaction. However, due to the complicated relationship between the deformation of soft tissue and the applied force in surgical simulation, it is a challenging task to achieve the balance between accuracy and efficiency. To improve the stability, continuity, and accuracy of flexible deformation and computational efficiency in haptic rendering, a novel approach based on spherical harmonic representations to simulating the virtual deformation of soft objects was proposed. The particular contents are as follows:Firstly, the method of haptic deformation modelling was studied detailedly, including the geometric modeling of soft object, expression of relationship between the deformation of soft tissue and the applied, control algorithm of haptic rendering and so on. A novel haptic deformation model based on spherical harmonic representation and principal components analysis was proposed. The object surface was represented by spherical harmonic, which features of orthonormality, rotational invariance, and multi-resolution. And, different spherical harmonic models were mapped into common reference system in order for corresponding coefficients in the different models to be comparable with method of principal components analysis and force feedback to be calculated by simplified deformation wave equation. Experiment results demonstrate that accurate flexible deformation is represented.Secondly, control algorithms of haptic rendering base on spherical harmonic were proposed. For example, in the field of collision detection, distance fields for rapid collision detection in haptic interaction based multi-resolution of spherical harmonic was proposed. In the field of collision response, given the density, Young’s modulus, Poisson ratio of the object and interactive position of the applied force, the haptic deformation model was estimated by a radial basis function based neural network model, which was trained with least error measure method. The simulation results indicated that the haptic deformation model could be accurately calculated using the proposed approach and the computational time was acceptable for real-time applications.Finally, experimental platform of haptic deformation rendering was established based vizard and sensable-phantom(?) desktopTM. Experiment results show that methods of haptic deformation modelling based on spherical harmonic and radial basis function neural network is suit for virtual reality applications for its interactive ability and realistic senses, such as virtual diagnosis, medical education, surgical training, medical database visualization and so on.