Abstract
An efficient stereo video coding method was proposed instead of calculating both disparity estimation (DE) and motion estimation (ME) of every stereo pairs. The first right frame was encoded firstly, and the first left frame was predicted by DE with adaptive windows. This first stereo pair was considered as I stereo pair. According to occlusion correlation between disparity vectors (DVs) and motion vectors (MVs), next three adjacent stereo pairs, which were referred to as P stereo pair, were put into content-based 4 dimensional units (4DUs). And MVs of the middle time blocks in 4DUs were referred to as the uniform MVs of the 4DUs, and were used to calculate the motion compensation. 4D matrix DCT (4D-MDCT) was used to eliminate spatial and temporal redundancy further. Experimental results show that the proposed scheme can reduce computational complexity more than 95% and 66% in DE and ME respectively by keeping image quality.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
