Frame rate up-conversion of compressed video using region segmentation and depth ordering

In this paper, a region-based method is proposed for frame rate up-conversion. To reduce spatial discontinuities and mitigate deformed structures commonly seen in block-based motion-compensated frame interpolation approaches, the proposed method interpolates frames based on depth-ordered moving regions. We develop a 3-parameter motion model to describe regions' moving trajectories, and use the local depth order among neighboring regions for occlusion during frame synthesis. Experimental results show that the proposed scheme is very effective in removing block artifacts in interpolated frames, and has a better visual quality compared to conventional methods.