Animating pictures with stochastic motion textures

Yung-Yu Chuang, Dan B Goldman, Ke Colin Zheng, Brian Curless, David H. Salesin, Richard Szeliski
2005 ACM Transactions on Graphics  
a) Japanese temple (b) Harbor (c) The Boat Studio (d) Argenteuil (e) Sunflower Figure 1 Sample input images we animate using our technique. The first two pictures are photographs of a Japanese Temple (a) and a harbor (b). The paintings shown in (c) and (d) are Claude Monet's Le bateau atelier (The Boat Studio) and The Bridge at Argenteuil. We also try our method on Van Gogh's Sunflower (e) to animate the flowers. Abstract In this paper, we explore the problem of taking a still picture and
more » ... it move in convincing ways. In this paper, we limit our domain to scenes containing passive elements that respond to natural forces in some oscillatory fashion. We use a semi-automatic approach, in which a human user segments the scene into a series of layers to be individually animated. The automatic part of the approach works by synthesizing a "stochastic motion texture" using a spectral method -i.e., a filtered noise spectrum whose inverse Fourier transform is the motion texture. The motion texture is a time-varying 2D displacement map, which is applied to each layer. The resulting warped layers are recomposited, along with "inpainting" to fill any holes, to form the animated frames. The result is a video texture created from a single still image, which has the advantages of being more controllable and of generally higher image quality and resolution than a video texture created from a video source. We demonstrate the technique on a variety of photographs and paintings.
doi:10.1145/1073204.1073273 fatcat:curogxgbuvbkdoepbielsfzn2i