基于嵌入式Linux的无线音视频通信系统的设计与实现

发布时间：2018-05-11 02:22

本文选题：无线音视频通信 + C/S　；参考：《杭州电子科技大学》2012年硕士论文

【摘要】：随着嵌入式技术和无线通信技术的迅速发展，特别是3G时代的到来和无线WiFi技术的广泛应用，为嵌入式无线音视频通信提供了有效的途径。然而对于普通用户来说，基于3G的可视电话给他们设定了两道门槛：一是高端的手机，，二是昂贵的资费。针对这一点，本文实现了基于WiFi技术和嵌入式Linux的低成本音视频通信方案。在此方案中，硬件平台采用S3C2440A ARM9开发板外接RT2571W WiFi模块；操作系统采用嵌入式Linux2.6.29；网络传输采用无线AP（Access Point）接入结合以太网传输的方式；通信系统采用C/S加P2P的网络架构，C/S架构下主要完成客户端与服务器之间各种信令的交互（例如注册、登录等信令及其应答），P2P架构下主要完成客户端与客户端之间音视频通信过程中语音和视频数据的点对点传输。本文的主要工作是设计和实现了音视频通信系统的架构和软件，包括服务器软件和客户端软件。服务器软件采用应用程序服务器与MySql数据库服务器相结合，提供客户端用户信息的验证、存储以及各种信令的应答、转发等功能。客户端软件是本文的工作重点，它是基于嵌入式Linux和Qt/Embedded开发的嵌入式软件，运用了音视频采集与压缩编码技术、同步技术以及网络传输技术，实现了音视频实时同步通信。主要研究工作归结为以下几点： 1、实现了基于Video4Linux2的视频采集，研究了MJPEG视频压缩算法，并根据前后帧图像变化检测的方法以及帧间压缩的思想，提出了基于变化区域检测提取的MJPEG改进算法，该改进算法在背景不变或存在背景短暂不变（比如视频通信时摄像头是静止的）的场合下能大大增加压缩率，减少压缩时间。 2、研究并实现了基于G.729B压缩算法的语音软件压缩方案和基于AMBE-2000压缩芯片的语音硬件压缩方案。对前者的工作有：实现在OSS音频编程接口上的语音数据的采集，研究了G.729B算法并对其标准C源码进行了优化来减少其算法复杂度。对后者的工作有：AMBE-2000语音模块的电路设计、驱动的编写以及相应应用程序的实现。比较并分析了上述两种方案，最终在客户端软件开发中采用了软件压缩方案。 3、从语音视频的采集、发送、接收和播放四个环节研究并实现了音视频同步技术。对网络传输技术做了研究，包括基于NAT穿透之UDP打洞技术实现了处于不同私网网段的客户端之间的P2P数据传输和自定义了方便简捷的通信协议。结合上述相关技术及Qt/Embedded图形开发框架实现了美观的可视化客户端软件。将客户端软件移植到ARM平台上，搭建了测试环境，最终实现了效果较好的无线音视频通信。
[Abstract]:With the rapid development of embedded technology and wireless communication technology, especially the arrival of 3G era and the wide application of wireless WiFi technology, it provides an effective way for embedded wireless audio and video communication. For ordinary users, however, 3G videophone sets two barriers: high-end phones and expensive tariffs. In view of this, this paper realizes the low cost audio and video communication scheme based on WiFi technology and embedded Linux. In this scheme, the hardware platform adopts the RT2571W WiFi module of S3C2440A ARM9 development board, the operating system adopts embedded Linux 2.6.29, the network transmission adopts wireless AP(Access point) access and Ethernet transmission. The communication system adopts the network architecture of C / S plus P2P. Under the architecture of C / S, the communication system mainly completes the interaction of various signaling between the client and the server (for example, registration, registration). Under the P2P architecture of signalling and answering such as login, the point-to-point transmission of voice and video data between client and client is mainly completed in the process of audio and video communication between client and client. The main work of this paper is to design and implement the architecture and software of audio and video communication system, including server software and client software. The server software combines the application server with the MySql database server to provide the functions of client user information verification, storage, response and forwarding of all kinds of signaling. Client software is the focus of this paper. It is an embedded software based on embedded Linux and Qt/Embedded. It uses audio and video acquisition and compression coding technology, synchronization technology and network transmission technology to realize real-time audio and video synchronous communication. The main research work can be summarized as follows: 1. The video acquisition based on Video4Linux2 is realized. The algorithm of MJPEG video compression is studied. According to the method of detecting the change of frame image and the idea of compression between frames, an improved MJPEG algorithm based on the detection of change region is proposed. The improved algorithm can greatly increase the compression ratio and reduce the compression time when the background is invariant or the background is temporary invariant (for example, the camera is still in video communication). 2. The speech software compression scheme based on G.729B compression algorithm and the speech hardware compression scheme based on AMBE-2000 compression chip are studied and implemented. The work of the former is as follows: to realize the acquisition of voice data on the OSS audio programming interface, the G.729B algorithm is studied and its standard C source code is optimized to reduce the complexity of the algorithm. The work of the latter includes the design of the circuit of the voice module: AMBE-2000, the design of the driver and the realization of the corresponding application program. The two schemes mentioned above are compared and analyzed. Finally, the software compression scheme is adopted in the client software development. 3. The audio and video synchronization technology is studied and realized from four links: audio and video acquisition, transmission, reception and playback. In this paper, the network transmission technology is studied, including UDP hole drilling technology based on NAT penetration to realize P2P data transmission between clients in different private network segments, and a convenient and simple communication protocol is defined. Combined with the above related technology and Qt/Embedded graphics development framework, the visual client software is realized. The client software is transplanted to the ARM platform, and the test environment is built. Finally, a better wireless audio and video communication is realized.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TN914;TP368.1

【参考文献】