Research On The Key Technology Of Video Compression Of The Railway Emergency Communication System

With the successful implementation of the railway “12th Five Year Plan”, the construction of Chinese high-speed railway is developing with a trend of fast increase.At the same time, the establishment of railway emergency rescue system is crucial to the rapid transportation safety. The railway emergency communication system can quickly establish the temporary rescue communication platform at the scene of an accident, which means it can provide a rich set of voice and video services for command and coordination. The Rescue command center could formulate an effective rescue plan accurately according to the real-time video information of the scene. It can be said that the video services played an important role in the emergency rescue system. So it is always a major bottleneck to keep a balance between coding efficiency and video quality. Based on the research of railway emergency communication system, this thesis mainly analyses the H.264 video compression coding technology. After analyzing the defects of the UMHexagonS algorithm and combing the macro-block mode selection strategy with the search template improvement strategy, the thesis chooses standard video test sequences for simulation and analysis.Firstly, based on the in-depth study of the system of railway emergency communication,the thesis compares MPEG-4 and H.264 video compression coding standards which are suitable for the system. According to the emergency service transmission protocol, it also carries out simulation analysis under the same conditions. Due to the railway emergency services’ particularity and complexity, this thesis chooses the H.264 technology as the railway emergency video compression standard and analyzes the main technologies of the H.264 which have advantages at their network adaption and anti-interference performance.Secondly, the thesis describes the principle of block matching theory and criteria in detail. It focuses on the classic motion estimation algorithm and analyzes the search principle and implementation process of FS, TSS, FSS, DS and HEXBS. In order to verify the performance and analyze the pros and cons of each algorithm, this thesis selects Foreman_cif test sequence to simulate, then compares the simulation results of TSS, FSS, DS, HEXBS with FS.Finally, according to the requirements of reliability and real-time in the video service of the railway emergency communication system, the thesis focuses on the theory and key techniques of UMHexagonS algorithm. After comparing with FS through simulation under the same conditions, it turns that the performance of UMHexagonS algorithm is much closer to the FS method. However, the combination of multiple search templates also makes the operation of the algorithm more complicated. So the UMHexagonS algorithm can be optimized from the analysis of macro block mode selection and improved search template.After combining all those strategies, the thesis selects three different sports level standard of test sequences to carry out the simulation. The results show that the improved UMHexagonS algorithm can well ensure the video quality and significantly improve the coding efficiency compared with the original UMHexagonS algorithm. This strategy effectively guarantees the reliability and instantaneity of the video services of the railway emergency communication.