论文部分内容阅读
Motion estimation algorithm has a high impact on the visual quality of an encoder for a given bit-rate, the aim of this thesis is to examine the role of motion estimation techniques in video compression and to investigate the performance of several motion estimation techniques widely accepted by various video coding standards for real time applications in cost effective manner.Motion estimation algorithm is regarded as an optimization problem in terms of its application and efficiency. Fast motion estimation techniques along with experimental results comparing their performances are presented in three directions: reducing computational complexity, true motion representation methods and improving coding efficiency. Recent advances in rate control techniques for video coding, such as new concepts in rate-distortion modeling and quality constrained control, are presented. With these techniques, the rate control performance can be improved.The paper not only summarizes these recent rate control techniques but also provides explicit directions for future rate control algorithm design and the gain in performance of PSNR with rate control in H.264 is also discussed.