Abstract

In a coding method which selects global motion compensation (GMC) or local motion compensation (LMC) as the motion compensation method for each block in an image, the appropriate selection of motion compensation methods is crucial for improving the coding efficiency. The difficulty of this GMC/LMC selection problem is mainly caused by the difference in the amount of motion information: GMC requires only a few motion parameters for the whole image, while LMC requires at least one motion vector per block. To develop an effective and practical selection criterion, we address the issue of: (1) estimating the number of bits used for the coding of the DCT coefficients from the mean absolute prediction error in a block, and (2) adopting a new function for evaluating the reconstruction error in a block and its side effect in future frames. Simulation results of the proposed coding method indicate that the coding efficiency is improved by up to 34.7% compared with conventional coding method which do not use GMC.

Full Text
Published version (Free)

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