Simulation of soft tissue using mass-spring model with simulated annealing optimization

Shaoping Xu,Peter Xiaoping Liu,Hua Zhang
DOI: https://doi.org/10.1109/ICAL.2009.5262704
2009-01-01
Abstract:A key challenge of the simulation of deformable soft tissue is to satisfy the conflicting requirements of real-time interactivity and physical realism. The behavior of soft tissue can be described by a mass-spring model provided that correct parameters, such as spring stiffness and viscosity, are used. In practice, such parameters are often determined by trial-and-error based on the visual effects of simulation. Therefore, it is very difficult to obtain accurate values, and the process is tedious and time consuming. In this paper, a heuristic optimization technique is proposed to identify these parameters of the mass-spring model for soft tissue simulation. We employ the simulated annealing algorithm to tune the parameters automatically until the deformation of simulated soft tissue approximates the reference one as defined by a new modified tensor-mass model in which we have taken into account the viscoelasticity of soft tissue. The proposed technique provides an automatic method to determine the parameters in mass-spring-based models.
What problem does this paper attempt to address?