| With the development of multimedia information technology, wireless multimedia sensor networks system had been widely applied, it consisted of video capture, video coding, video transmission, video decoding and play. Video transmission technology and video compression technology were often utilized in various fields, such as video surveillance, digital TV, remote video conference and so on. Due to the limitation of communication channel bandwidth in wireless video transmission, it was necessary to control the sending ratio of videos, this was called rate control technology which was aim to to make sure the smooth transmission of rate stream and the rational use of channel bandwidth in wireless video transmission. It could control video compression ratio and guarantee playback quality.The article topic selected from wireless multimedia sensor network program in Chinese Academy of Shenzhen Institute of Advanced Technology, focusing on video coding and rate control. First, the paper introduced the key technologies in the system which consisted of wireless video transmission and H.264 video compression and rate control. Do more depth research on rate control part by taking advantage of the next-generation video compression standard H.264 protocol; then according to the open encoder X264 to compress videos. Among them, X264 was changed from H.264 and very practical, X264 included in several rate control algorithms: CRF, CQP,VBV-CBR, ABR and so on.First, the title tested and analysed the existing several rate control methods ‘s performance of controlling rate, and compared VBV-CBR to ABR, selected ABR algorithm to do deeper research and improvement.Second, according to ABR experiment result, namely between the target bit-rate and the actual bit-rate’s deviation was too large which led to the imprecise and uneven of allocating bit-rate, A novel rate control algorithm had been proposed to improve the efficiency of rate control. The paper had taken the following two considerations: the significant change in the difference value of the actual bits and expected bits in the last encoded frame which changed obviously with the difference of the contents and velocity in pictures, then adjusting the QP in different degree; at the same time,according to the difference value between the actual bit-rate and target bit-rate of encoded frames to adjust bit-rate.Third, according to the the executing platform of algorithm, we executed the X264 open source code to unpack, compile in PC Linux system; then made use of the C to develop and modify the part of rate control; coding, decoding and play the YUV standard test videos. Compared to original algorithm, experiments showed that the output bit-rate was more closely to expected bit-rate in the suggested algorithm without losing video quality, it saved more bit-rates and was more suitable for video transmission.Moreover, the paper established a wireless multimedia sensor network platform which realized to make use of V4L2 interface to achieve capture of two videos, then crossing compiled x264 source code in PC, then compressed the capture videos and sent the compressed video data by RTP transmission protocol in ARM9; according to FFMPEG source code and MFC interface to accomplish the video decoding and play part, the MFC interface player can check the quality of video, the system can provide a feasible solution scheme of wireless video compress transmission. |
PDF Full Text Request