面向家居安全的移动视频监控识别系统的研究与实现

发布时间：2018-09-18 09:46

【摘要】：随着家居安全意识的不断提高,室内移动智能视频监控技术在安全应用领域中扮演着重要的角色,其中以人为目标实时有效地监控检测识别室内异常行为一直是研究的难点。由于智能终端的便捷性,移动性和集成性,将本文提出家居安全的异常行为检测算法移植到智能手机平台上,将改变传统的监控视频处理方式,使其更具灵活性,提高了视频监控识别系统的实时性。同时综合利用移动互联网和互联网技术来完善视频监控识别系统的报警功能。本文通过分析面向家居安全的移动视频监控识别系统的功能需求,主要是从家居安全的异常行为检测算法、面向智能手机的异常行为视觉检测识别和基于移动互联网和互联网技术的报警通信三个方面进行研究与实现。论文的主要工作如下:首先,针对常见的两种家居安全的异常行为即异常入室行为和室内独守老人摔倒行为,我们提出了一种家居安全的异常行为检测算法。该算法把这两种异常行为的检测识别方法融合在一起,减少了检测算法的冗余度和复杂度。对于异常入室行为的检测识别是通过对检测到的运动目标大小进行合理性判别。而对于摔倒行为检测是在确定异常运动目标,宽长比的基础上,采用有效面积、重心变化率和状态维持机制对检测结果排错修正,从而提高了检测系统的可靠性。家居安全的异常行为检测算法的复杂度比较低,效率高,可靠性高,具有较好的自适应性。其次,为了充分合理利用智能手机集成的硬件资源,进一步提高室内异常行为的检测识别的实时性,移动性和便捷性,我们对面向智能手机的异常行为视觉检测系统进行了深入研究,并且在智能手机平台上实现原始视频图像数据采集,家居安全的异常行为检测算法,结果图像显示以及系统报警机制。面向智能手机的异常行为视觉检测系统避免了传统的视频监控行为检测系统在安装成本,便捷性和实时性等方面的约束。最后,系统改进了家居安全的异常行为检测算法和单一的SMS(Short Messaging Service)报警机制,并且将家居安全的异常行为检测,用户查看报警信息等功能整合在同一个应用程序系统中,从而最终实现了面向家居安全的移动视频监控识别系统。在系统优化中,改进型家居安全的异常行为检算法综合利用了智能手机的光环境传感器硬件资源,提高了系统中异常行为检测对光照强度变化的抗干扰性;改进型报警机制将基于移动互联网/互联网技术的主动型SMS报警方式和被动型图文报警方式相结合,完善了系统报警机制的功能。
[Abstract]:With the continuous improvement of home safety awareness, indoor mobile intelligent video surveillance technology plays an important role in the field of security applications. Because of the convenience, mobility and integration of intelligent terminals, this paper proposes an anomaly detection algorithm for home security to be transplanted to smart phone platform, which will change the traditional video processing mode and make it more flexible. The real-time performance of video surveillance and recognition system is improved. At the same time, the mobile Internet and Internet technology are used to improve the alarm function of video surveillance and identification system. This paper analyzes the functional requirements of the mobile video surveillance and identification system for home security, mainly from the home security abnormal behavior detection algorithm. The research and implementation of smart phone oriented vision detection and recognition of abnormal behavior and alarm communication based on mobile Internet and Internet technology are carried out. The main work of this paper is as follows: firstly, aiming at two kinds of abnormal behavior of home safety, namely, abnormal behavior of entering the house and falling down behavior of the elderly alone in the room, we propose a detection algorithm of abnormal behavior of home safety. The algorithm combines the two detection and recognition methods of abnormal behavior and reduces the redundancy and complexity of the detection algorithm. The detection and recognition of abnormal entry behavior is based on the rationality of the size of the detected moving object. On the basis of determining the target of abnormal motion and the ratio of width to length, the effective area, the change rate of gravity center and the state maintenance mechanism are adopted to correct the detection results, thus improving the reliability of the detection system. The anomaly detection algorithm of home safety has low complexity, high efficiency, high reliability and good adaptability. Secondly, in order to make full use of the integrated hardware resources of smart phones and further improve the real-time, mobility and convenience of indoor abnormal behavior detection and identification, We have carried on the thorough research to the smart phone oriented abnormal behavior vision detection system, and realized the original video image data acquisition, the home security unusual behavior detection algorithm on the smart phone platform, Results Image display and system alarm mechanism. The smart phone oriented visual detection system for abnormal behavior avoids the constraints of the traditional video surveillance behavior detection system in terms of installation cost convenience and real-time. Finally, the system improves the abnormal behavior detection algorithm of home safety and the single SMS (Short Messaging Service) alarm mechanism, and integrates the abnormal behavior detection of home safety, the user view alarm information and other functions in the same application system. Finally, the mobile video surveillance and identification system for home security is realized. In the system optimization, the improved abnormal behavior detection algorithm of home safety synthetically utilizes the hardware resources of the light environment sensor of the smart phone, and improves the anti-interference ability of the abnormal behavior detection to the illumination intensity change in the system. The improved alarm mechanism combines the active SMS alarm mode based on the mobile Internet / Internet technology and the passive graphic and text alarm mode to perfect the function of the system alarm mechanism.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN948.6

【参考文献】