A real-time 2D to 3D video conversion algorithm based on image shear transformation

3D movies can bring immersive feelings to people. They are popular nowadays. At present, numbers of 3D movies are produced each year, but these can not satisfy the growing demands. The productions of 3D movies are complex, and which cost high and takes a long time. 2D to 3D conversion technologies can convert the existing 2D video to 3D video quickly and effectively, which could reduce the difficulty and cost. Therefore, these technologies can alleviate the problem of insufficient 3D videos to a great degree. A real-time 2D to 3D algorithm based on image shear transformation is presented in this paper. According to the shear transformation factor, each frame in the 2D video is converted along the positive and the negative direction of the x axis respectively to obtained the images of the left and the right eyes. A mathematical model of shear transformation factor and stereo view depth is also discussed, therefore to control stereo effect quantitatively. The computational complexity is reduced in this algorithm to meet real time requirements, and the 3D effect can be adjusted to obtain good results. The algorithm is used to produce several 3D videos. The experimental results show that a satisfactory 3D effect can be got automatically and in real time