Hierarchical Retargetting of 2D Motion Fields to the Animation of 3D Plant Models

Abstract

The complexity of animating trees, shrubs and foliage is an impediment to the efficient and realistic depiction of natural environments. This paper presents an algorithm to extract, from a single video sequence, motion fields of real shrubs under the influence of wind, and to transfer this motion to the animation of complex, synthetic 3D plant models. The extracted motion is retargeted without requiring physical simulation. First, feature tracking is applied to the video footage, allowing the 2D position and velocity of automatically identified features to be clustered. A key contribution of the method is that the hierarchy obtained through statistical clustering can be used to synthesize a 2D hierarchical geometric structure of branches that terminates according to the cut-off threshold of a classification algorithm. This step extracts both the shape and the motion of a hierarchy of features groups that are identified as geometrical branches. The 2D hierarchy is then extended to three dimensions using the estimated spatial distribution of the features within each group. Another key contribution is that this 3D hierarchical structure can be efficiently used as a motion controller to animate any complex 3D model of similar but non-identical plants using a standard skinning algorithm. Thus, a single video source of a moving shrub becomes an input device for a large class of virtual shrubs. We illustrate the results on two examples of shrubs and one outdoor tree. Extensions to other outdoor plants are discussed.