A New Video Coding Approach Based on Object- Extraction

This paper describes efficient object-based video coding schemes suitable for content-based multimedia streaming systems. Anew two video coding approaches are presented to study the effect of the object based motion estimation in video coding and the effect of the object based video coding on the compression quality, respectively. The first approach is based on a new motion estimation technique; based on arbitrary shapedregions. The second approach is based on the video object extraction and a new motion estimation technique based on arbitrary shaped-regions. The proposed methods are applied on videos containing a variety of scenarios such as multiple objects undergoing occlusion, splitting, merging, entering and exiting, as well as a changing background. The simulation results are introduced in a comparison form with some of recent video coding method. For all the videos, the proposed approach displays higher Peak Signal to Noise Ratio (PSNR) compared to other methods, and provides comparable or better compression than some of recent video coding techniques. Index Term— Video coding; image segmentation; object extraction; MPEG-4; video object planes (VOPs); motion vectors estimation