【作者】 宁以风；

【导师】 孙涌；

【作者基本信息】 苏州大学 ， 计算机应用技术， 2007， 硕士

【摘要】 随着下一代网络技术的逐步推进,Web服务技术的深化,网络摄像机的智能化发展,网络技术、多媒体技术、异构平台的软件支撑技术日益成熟。语音、视频、数据等多种业务能够融合在一个综合性的、全开放的网络平台上。实时动态网络视频的应用系统是通讯技术、智能控制技术和计算机技术的有机结合,将被广泛应用于安全防范、远程监控、远程教学、病房监护、社区服务等各计算机应用领域。本文在网络视频技术和Internet技术的基础上,提出多用户远程访问、控制实时动态视频图像应用系统的设计方法,构建多用户实时动态网络视频控制系统的实现方案。在系统的分析、设计中,运用面向对象软件工程方法(OOSE方法),使用统一建模语言(UML),对系统进行了面向对象的分析和设计;运用基于形式的、数学的规约技术,采用形式化语言——Z规约对构造系统的关键模块进行了探讨;从数据封装、抽象数据类型、信息隐藏等技术角度,对系统核心算法进行了解析。在系统核心算法的处理中,应用运筹学中排队论的思想,解决多用户对硬件资源的竞争问题;分析系统算法的性能,提出实施系统优化设计的策略,并给出一个有效的处理最典型的面部检测的新算法。更多还原

【Abstract】 With the development of the technology of NGN(Next Generation Network), Web Service, intelligent network Camera, the technology of support for network, multimedia and heterogeneous system has been more and more practicable. So many businesses, such as voice, video, dates, and so on, can be integrated with a synthesized, full-opened network platform. The system of real-time and dynamic network video, which integrates communication technology, intelligent control technology, will be used widely in all field of computer application, such as safety precautions, long-range surveillance and control, teleclass, sickroom electronic surveillance, community service, and so on.The paper, based on the thesis of internet and network video, designs a scheme for multi-users access real-time and dynamic video system, and puts forward a blue-print for it. The paper applies OOSE, UML technology to design and implementation of the system. It discusses how to use the formal mathematic specifications to design key-block of system. It researches on implementation of system combined with data encapsulation, abstract data type, information hiding. In implementation of the core algorithm organization, the paper applies the thesis of queue in operational research to solve the competition of hardware when multi-users access. Also, after analyzing the performance of system, the paper puts forward a strategy for the optimization of system. Besides, one novel efficient face detection and tracking algorithm are presented, which adapts for the most typical personal secret controlling.更多还原