Video Coding And Decoding Technology Research Based On Embedded Linux

With the rapid development of multimedia information technology, the demandsfor the video quality are also increasing rapidly in both optoelectronic measurementsystems and video surveillance systems. The optoelectronic system has evolved into alarge distributed multi-node system. For the network video surveillance, remotevideos need to be aggregated to the command center in real time, and a large amountsof data need to be transmitted in the limited bandwidth. So, high-speed and real-timevideo compression and transmission become the hotspot and difficulty. On the onehand, the high-resolution images are widely used to improve the measurementaccuracy in the optoelectronic systems, real-time remote transmission is the key; Onthe other hand, the transmission capacity of the network is severely limited inlow-bandwidth scenarios, the efficient image compression algorithm required forthe rate control. In addition, the devices in optoelectronic measurement, includingvideo compression device and transmission system, should keep small volume andlow power consumption for mobility. Response to these problems, this paper expandsthe in-depth study of the real-time video coding and decoding around the H.264videocompression algorithm.In order to build a small, and low-power system, the article conducted theresearch to the current video coding and decoding platform, and then chose the H.264Codec--MG3500ASIC platform as the hardware of the SoC. At the same time, thearticle transplanted the embedded Linux OS, built the software development platform.The experimental results show that the system meets the coding and decodingrequirements of PAL(704X576), NTSC(720X480),720p60,1080i60,1080p30et.al.There is a high demand for the real-time video in the optoelectronicmeasurement system, especially for the high definition video. In order to transport thevideo in the network, the article studied the H.264algorithm, and proposed a low-latency optimization algorithm, which considered the whole delay in imagecapturing, coding, transmission and decoding. On the basis of the rate control, it ispossible to control the delay effectively. In order to achieve a one-to-many networktransmission, this article uses Multicast and RTSP protocol to improve thetransmission efficiency, reduce network bandwidth. The delay is below300ms.This paper presents a low bitrate H.264rate control algorithm for the needs ofaudio and video communication under the low bandwidth3G mobile network. Theexperiments show that the algorithm can effectively control the encoding bitrate toreduce network bandwidth requirements such as CIF at160kb/s, while the imagequality is within the allowable range. The system completed the AAC audio codingalgorithm and the synchronization of audio and video in decoding. Meanwhile, thearticle designed to achieve the audio and video transmission at a dual-link loadbalancing network, which can expand the3G network bandwidth. In this paper, FECalgorithm to achieve a network packet loss recovery, improved the quality of audioand video transmission.In addition, in order to achieve the real-time transmission of H.264stream in thetelemetry channel, the paper also studied PCM coding algorithm and designed anPCM encoder combined ASIC and FPGA, which can encode the H.264stream to aserial data under the PCM frame format and send it in the telemetry channel.