A comparative study of physics engines for modeling soft tissue deformation

Deformation of objects geometry in computer graphics has been based on a number of different approaches. Two classes of methods are highlighted: physics-based and geometry-based. Physics-based methods, such as Mass-Spring Damper, Finite Element Method and Point-Associated Finite Field have been applied to model soft tissue. However, they all have difficulties in handling the compromise between physical realism and performance. Such difficulties can comprehend the complexity in setting physical parameters of the object, large computational cost and high complexity in implementation. This work proposes a case study evaluation on the Mass-Spring Damper model. A set of well known physical engines are analyzed, allowing to understand how the implementation decisions influence performance and accuracy. These results are compared with an implementation especially proposed for this study.