Research On Time-performance Optimization Of Software Encode Topology Based On H.264

With the extensive development of the applications of video surveillance system, network video monitoring system has become the mainstream. For the requirements of transmitting the video image highly active under a limited bandwidth, the JVT (Joint Vedio Team) in 2003 proposed a new generation video compression standard H.264. And compared to the previous video compression standards, H.264 video compression standard provides higher efficiency, greater error recovery, but increased the complexity of encryption algorithm, which will inevitably lead to additional consumption of encoding time, at the same time limits its application in real-time monitoring system. To resolve the problem of the additional time-performance consumption of H.264 software encoding, the paper firstly analyses the information source model of H.264, and presents an adaptive time-performance optimization model based on progressive-encode and evaluation algorithm of image quality. In this model, firstly according to the current software environment set the mode to organize the original image, secondly do an accessment to the new image by the evaluation algorithm of image quality, which helps to decide the selected model is reasonable; Finally, using the image which gotten by preprocessing the original image with current selected mode as the source data of H.264 encoder.On the basis of this model, we carried out the design and implementation of real-time video encoding system based on H.264, its main functions including automatically detecting the software environment of the system, preprocessing the original image, intelligently evaluating the quality of image, H.264 video encoding, encoding time performance analysis, the paper provides the deliberate design of system structure and module dividing, and the critical technique, deal process, as well as core data structure are described in detail in the core modules. The experimental results show that the real-time encoding system based on this optimization model, under the premise of assured image quality, reduce encoding time consuming to enhance the encoder's real-time ability obviously.