New MCTF method based on adaptive direction of motion vector for interframe wavelet coding

With recent growth of multimedia applications for delivering compressed bitstream over networks, scalability has become an important feature in video coding. The interframe wavelet coding (IWC) inherently provides temporal resolution, spatial resolution, and SNR scalability through wavelet transform and embedded quantization. In this paper, we propose a new motion compensated temporal filtering (MCTF) method. This method adaptively changes the directions for motion compensation in a group-of-pictures (GOP). Three modes are selected for the directions of motion estimation based on a criterion using motion information at boundary area in image sequence. This proposed method is composed of two parts. The first part is the extraction of motion information at boundary area. The second part is to process temporal filtering according to the selected direction for motion estimation. Simulation results show that the proposed method outperforms the conventional IWC method by 0.4 1.0 dB.