Animating Animal Motion From Still

 

 

 

 

 

 

 

 

Text Box: The first step in the algorithm is extracting the snapshots from the still images. The second step is building a shape context descriptor that should exhaust all the information in the snapshots. The third step is finding the distances between the snapshots, giving us a NxN matrix which contains the distance between any two snapshots. The fourth step is finding the optimal order between the snapshots that best represents the animal's motion path. The Fifth step is refining the snapshots so that they all have a similar color and texture and similar pose relative the camera. The final two steps create the animation. This is done by morphing between any two snapshots and blending the morph result on the background image.   In this work we attempt to implement the article "Animating animal motion from still" (Liu 2005). The goal of this project is to build a tool that creates an animation of animal motion from a still picture that includes a couple snapshots of the animal.  

Abstract

 

Abstract

Text Box: In this work we attempt to implement the article "Animating Animal Motion From Still" (Liu 2005). The goal of this project is to build a tool that creates an animation of animal motion from a still picture that includes a couple snapshots of the animal.

 

 

 

 

 

 

 

 

H:\semester7\Flock\ProjectBook\Presentation\Siggraph_asia_2008.JPG