Motion-Compensated Temporal Filtering For Scalable Video Coding

At present, by scalable coding, just a one-time code can be to meet different bandwidth, different resolution and frame rate applications. Among them, scalable video coding time-domain scalability using motion compensated temporal filtering (MCTF, Motion-Compensated Temporal Filtering) technology to a number of consecutive video frame image group composed of a wavelet decomposition in order to provide time-domain can be divided into Class of. The MCTF the input of the original image group (Group of pciuters, GOP) through the process of lifting wavelet projections are high-pass image, the update process by upgrading the wavelet to be low-pass image, and then low-pass image Repeat this process until it reaches the required decomposition the decomposition series. Due to the high-pass and low-pass images related to each other very small, so they were various independent scalable coding is possible. However, the very nature of MCTF algorithm to determine its encoding architecture, the complexity of large, so the succession of his ideas to improve their algorithm on the basis of reducing its complexity as a problem to be solved.This paper proposes a motion compensated temporal filtering to improve the motion search algorithm. The main objective is to further the coding efficiency and low-pass images to achieve a balance between visual quality, and improved the search algorithm is applied to the update step constraint conditions, it mainly thinking of using backtracking algorithm, by searching for updates pixels in the corresponding block range of motion vector prediction based on the overlap part of the overall motion vectors to determine exactly the same, and then decide whether to abandon the update operation. This retains the benefits of binding updates, the image of the subjective quality can be guaranteed, while further reducing the computational complexity. Experimental results show that the improved algorithm not only will not affect the earlier study, the introduction of MCTF update the constraints on low-pass image subjective quality improvement and coding efficiency will also be a certain margin improved.