Visual perception theory guided depth motion estimation

Bing Li,De Xu,Songhe Feng,Fangshi Wang
DOI: https://doi.org/10.1007/978-3-540-69423-6_20
2007-01-01
Abstract:Motion estimation is an important and computationally intensive task in video coding and video analysis. But existent motion estimation algorithms mainly focus on 2-D image plane motion and neglect the motion in depth direction, which we call it depth motion in this paper. There are even few researches on the depth motion, their methods are complex and most of them need binocular images. In this work, visual perception theory is used to estimate the depth motion. A novel depth motion estimate method is proposed base on visual perception theory and it can estimate the depth motion from just monocular video. Experimental results show that our model is simple, effective and corresponds to the human perception.
What problem does this paper attempt to address?