【作者】 张鹏；

【导师】 贾敏智；

【作者基本信息】 太原理工大学 ， 控制理论与控制工程， 2008， 硕士

【摘要】 随着信息技术和计算机互联网技术的飞速发展,网络化的视频监控技术已经广泛的得到了人们的关注,并得到了广泛的应用。凡是网络能够到达的地方就能通过网络视频监控技术进行远程的视频监控。视频信息是人类获取信息的最主要载体,而人类获取信息中70%来自于视觉。由于视频信息经数字化处理后具有很强的抗干扰能力、易于加密等优点,因此在监控方面得到了广泛的应用。但是由于数字化的视频信息对存储和传输有很高的要求,因此视频信息的压缩编码对存储和传输视频信息具有重要的意义。而视频信息的压缩以及压缩后视频信息的传输是视频监控的核心技术,目前压缩视频的编码方法主要有:H.263、H.264、以及MPEG-4等,本文采用的是MPEG-4视频编码。同时采用RTP/RTCP实时传输协议对视频流进行网络传输。本文在嵌入式Linux的ARM9平台上实现网络视频监控,主要完成的是服务器端的各部分功能。流程大致如下:首先利用USB摄像头采集数字图像,然后把采集到的视频信息传输到具有ARM920T核的S3C2410芯片进行MPEG-4视频压缩,接着把压缩后的视频流打包成RTP数据包进行网络传输。主要的工作有嵌入式平台的建立以及各个功能模块的程序实现。而功能模块主要包括视频采集模块、视频编码模块、视频传输模块。嵌入式系统平台的构建主要包括系统硬件平台的设计和系统软件平台的搭建。硬件平台的设计主要有:了解处理器S3C2410的体系结构,USB接口电路的设计,网络接口的设计,存储器电路的设计,电源部分的设计等;系统软件平台的搭建主要包括:宿主机交叉编译环境的建立;BootLoader的移植;嵌入式Linux内核的裁剪与移植以及yaffs文件系统的建立。在视频采集模块中,首先在Linux内核加载摄像头芯片的驱动和Linux的应用程序接口Video4Linux;然后利用Video4Linux,通过USB摄像头进行视频采集。在视频编码模块中,采用的是最新的MPEG-4codec—XVID进行视频编码的。主要是把XVID生成静态链接库,然后移植到ARM中,通过编写应用程序实现对视频信号的MPEG-4视频编码。在视频传输模块中,首先是对编码的MPEG-4码流进行打包,然后利用多媒体的实时传输协议RTP和实时传输控制协议RTCP对视频流进行网络传输。更多还原

【Abstract】 With the rapid development of information technologies and computer network technology, remote video surveillance based on network is in extensive consideration of people, and have been widely used. With this technology, anywhere a network is connected, a user can acquire video data and monitor the status of equipments in long distance.As is known to all, video information is the main carrier that the people get the information. And 70%’s information that the people get was from the video information. As the digital signal process of the video information, the information will have many merits, such as good anti-jamming, facility encryption. Therefore the video information has been widely used in video surveillance. However, because of the video information have a high demand in storage and information transmission, Video compression play a good part on it. And video compression and transmission of compressed video information are the core technology of video surveillance. Currently, the main video encode method have H.263, H.264 and MPEG-4, etc. This paper is based on MPEG-4 video encode, and using RTP / RTCP Real-time Transport Protocol to transmit the video streaming on network.In this paper, we realize the function of the network video surveillance on service side, base on the embedded Linux system on ARM9 platform. The main process is as follow: firstly, use the USB digital camera to capture the image, and then put the video information into the S3C2410 microcontroller, and then encode the video information; at last, package the compressed video information into RTP packet, and then translate into the Internet. this paper have built the platform for embedded system development and realized three main function modules which include the video capture module, video encode module and the video transmission module.Construct the embedded system platform include designing of the hardware platform and building the system software platform. Hardware platform design includes: understand the S3C2410 microcontroller architecture, USB interface circuit design, network interface design, the memory circuit design, power design. And the system software platform includes: establish the cross-compiler environment on the host machine, transplant the BootLoader and the Linux kernel on the ARM, build the yaffs file system.In the video capture modules, Firstly, loader the camera chip driver and the Linux Application Programming Interface-Video4Linux in the Linux kernel, and then we use the USB camera to capture video data with the Video4Linux.In the video coding module, we use the latest MPEG-4 codec—XVID to encode video data. mainly to generate the static link library for XVID, and then transplanted to ARM, and then through writing applications to achieve the MPEG-4 video coding for video signals.In the video transmission module, firstly to encode the MPEG-4 package streams, and then use the real-time transmission protocol RTP of multimedia and real-time transmission control protocol RTCP to transmit the video stream.更多还原