2D-to-3D conversion based on depth-from-motion

With the development of computer vision technology, stereo video technology gets more and more popular. However, compared with the growing number of stereoscopic display devices, the 3D content for display is obviously insufficient. On the other hand, there exists tremendous amount of 2D video resources. As a result, the 2D to 3D video conversion becomes a common interest of both academic and industrial communities. This paper focuses on the characteristics of 2D to 3D technology, summarizes different transformation methods, and presents a semi-automatic depth extraction scheme from motion estimation. The experimental results show that this scheme can obtain 3D depth maps with relatively high quality.