Research And Implementation Of Rate Distortion Optimization Algorithm Based On Source Distortion

with people’s growing demand for video image quality, there is an urgent need for efficient video compression coding technology. To save more transmission bit, achieve better video quality and improve the encoder performance under the particular encoding standard, the optimization technology has become an important part of the video image compression technology. rate distortion optimization technique is the most commonly used optimization technique. we can achieve an optimal balance between the rate and distortion and obtain a better coding performance by using the rate distortion optimization technique.This thesis systematically introduced the rate distortion theory, rate distortion optimization techniques and rate distortion models, besides, the independent coding rate distortion optimization technology and global rate distortion optimization technique were analyzed. According to the time-domain between the source video image frames, we proposed a time-domain global rate distortion optimization technique. Based on this technology, the quantization parameter adaptive adjustment algorithm and Lagrangian adaptive adjusting algorithm were proposed. The main work of this thesis has three parts.1. Proposed a new distortion model to describe the distortion, based on the quantization distortion and motion compensation distortion, the source distortion is proposed.This thesis analyzed the existing distortion calculation method and founds it will lead to error, so we proposed a improved distortion estimation method. Besides, we established a distortion model to describe the relation among three kinds of distortion.2. Proposed a quantization parameter adaptive adjustment algorithm, established a source distortion propagation chain to estimate the exercise intensity of the encoding unit, and adaptively adjusted the quantization parameter based on the calculated exercise intensity. This algorithm includes frame level quantization parameter adaptive adjustment algorithm and GOP level quantization parameter adaptive adjusting algorithm. The idea of the layer coding has been applied to the GOP level quantization parameter adaptive adjustment algorithm.3. Proposed a Lagrangian multiplier adjustment algorithm, because the complexity of the global rate distortion optimization is high, this thesis presented a time-domain rate distortion optimization model. Based on the distortion model, we established the timedomain propagation distortion model. Propagation factor was introduced to adjust the Lagrange multipliers. When the quantization parameter adaptive adjustment algorithm and Lagrange multipliers adjustment algorithm both opened, the joint adaptive adjusting algorithm is formed.Through researching the above three aspects, rate distortion optimization algorithm based on the source distortion is carried out on Chinese audio and video standard AVS reference model software RD12.0 and compared with the general test conditions CTC.From the test results, it can be seen that the algorithm proposed in this paper can obtain larger BD-Rate gain and greatly improve the coding performance. The work of this thesis has been adopted by the AVS group and integrated into the reference model software RD as a coding optimization tool.