Research Of Motion Estimation Algorithm Based On H.264/AVC Video Codec

The video coding standard H.264 is developed by the International Telecommunication Union (ITU) and Moving Picture Experts Group (MPEG) formed the Joint Expert Group (JVT). This coding standard has high coding efficiency for low-bandwidth, high-quality network video applications. However, in order to improve the coding efficiency, H.264 using a number of high computational complexities of the algorithm. Motion estimation is the core video compression algorithm in coding system, accounting about 60% computation. Therefore, how to improve the efficiency of motion estimation algorithm to make the search process more robust, faster, more efficient become one of the hotspots.Through the in-depth study of classic and traditional fast motion estimation algorithms, in this paper, we proposed two algorithms upon traditional algorithms: Improvement Based on Adaptive Flat Search Template of ARPS and Fast Mode Selection Algorithm Based on Image Intensity Classification.In the Improvement Based on Adaptive Flat Search Template of ARPS, the algorithm puts forward new ideas about the initial prediction motion vector algorithm, early termination criteria and through the trend of the object's different sports using different search templates. Experiments show that the proposed algorithm has higher speed than the original ARPS algorithm under the premise of ensure image quality.In the Fast Mode Selection Algorithm Based on Image Intensity Classification, the algorithm build a classification threshold based on the movement intensity of image, through the level of the image to use the corresponding prediction model to quickly select the best prediction model. Experiments show that the algorithm proposed in this paper has substantial increase in coding efficiency under the premise of maintain the basic rate-distortion characteristics.