For Three-dimensional Video Quality Assessment And Rate Control

As the rapid development of information technology, the 2D visualization technology based on flat-image description could not satisfy the increasing needs that people required more vivid and unartificial scene. Stereo video was a new-style video with third dimension and interaction function, which could meet the user's operation to audio-visual object from various angles through multiple points of video data. And also it provided the interaction capability of roaming scenes and embodied the direction of cyberization, interaction and third dimension to the next generation of multimedia applications. In all, stereo video had a golden prospect in various aspects. While it had a series of advantages, stereo video brought video coding data dramatic increased, which made the data storage and transmission became very difficult. How to guarantee the quality of the video, and at the same time to maintain a relatively high compression ratio and stable bit rate, that was a key technology to stereo video. Hence, it has great practical significance to establish a new high-powered method for stereo video coding based on the existing video coding schemes.This paper is organized as follows:1. A new algorithm for stereo video image quality assessment is proposed based on the basic characteristics of human visual system (HVS) such as visual non-linearity, contrast sensitivity, multi-channel model and so on.2. In view of the drawbacks of TM5 rate-control algorithm such as poor rate-control precision, high computational complexity due to macroblock-level adaptive quantification and so on, a novel TM5-based frame-level bit-allocation and rate-control algorithm is proposed. In the proposed algorithm, the frame-level target bit-allocation method is improved and a new rate-distortion model is incorporated to implement rate-prediction.3. In this section, based on the current research findings of visual physiology and psychology, the main characteristics of HVS were got analyzed and improved. Then the classical rate control algorithm of JVT-G012 was improved by incorporating the HVS properties to improve the subjective image quality under a given target bit-rate. Finally, a novel macro-block rate control algorithm combined with visual perception was proposed.4. For the technology of stereo video coding, that were introduced briefly about the research status and key technologies firstly. Secondly, based on the research of image quality assessment and rate control aboved, the coding structure of single point was modified to meet stereo video. And then the rate control algorithm for stereo video coding was studied mostly. Finally, experiments were conducted using the proposed coding scheme.