Dynamic Texture Segmentation Using Fourier Transform
Jianghong Li, Liang Chen, Yuanhu Cai · Modern Applied Science · 2009
Dynamic textures are temporally continuous and infinitely varying sequences of images with certain spatial and temporal stationarity properties and have many potential applications such as computer graphics, computer vision, animation, and computer games. These applications usually require dynamic texture segmentation that is not an easy task especially when the background is cluttered and textured. In this paper, we analyze the spatio-temporal slices of dynamic texture using Fourier analysis and then we propose a simple and efficient approach to dynamic texture segmentation using 3D Fourier transform. Experiments on several dynamic texture sequences show the effectiveness of the proposed approach.