A coder for affine transformation in H.263

Coding for video transferring in communication networks is essential because the digital representation of image sequences requires a very large number of bits. Similarity in image sequences causes transferring whole of image data to be unnecessary. The statistical analysis of video signals indicates that there is a strong correlation between successive picture frames as well as within the picture elements themselves. Many methods and coders to find this information have been proposed up to now. In this research, we propose a new coder using Affine Transformation along with BMA technique for motion compensation in H.263 standard. In order to decrease the amount of computations in online communications, we have utilized new conditional motion estimation. Simulation results clearly prove that our method reduces bit-rate with respect to H.263 standard.