基于ARM11的移动物体监控系统的设计与研究

发布时间：2018-04-05 14:14

本文选题：移动物体监控　切入点：ARM11　出处：《西安建筑科技大学》2017年硕士论文

【摘要】：近年来,随着社会的进步和科技的飞速发展,移动物体监控技术也得到了迅速发展。移动物体监控在保护人身安全和财产安全中发挥着越来越重要的作用。然而传统的视频监控大多只是视频的录制,不能对移动物体进行提取,后续工作是需要人工进行实时观看处理,效率过低且费时。本设计采用移动物体检测技术和视频监控技术相结合的方法,实现了基于ARM11的移动物体监控系统的设计。基于对监控的基本功能和性能的考虑,设计了声音报警子系统和摄像头开发子系统,并完成了从web浏览器访问web服务器中保存的移动物体图片和视频的设计。软件设计的主要任务是对移动物体监控系统提出一种新的移动物体检测算法。首先分析了移动物体检测的研究现状、移动物体检测算法和视频监控中存在的主要问题。在对时间平均法和像素估计法进行建模分析后,发现时间平均法得到的背景模型中存在虚影;像素估计法得到的背景图像会将物体的倒影误判为前景,出现噪声导致检测效果不准确。针对时间平均法和像素估计法的自身的不足,本文将两种算法的优势进行融合,提出了一种改进像素估计法,使两种算法达到了优势互补的效果,改进像素估计法得到的背景图像相对稳定,而且可以提取到理想的前景目标。具有较好的鲁棒性和自适应性,为移动物体监控提出了一种有效的算法。最后对基于ARM11的移动物体监控系统作了测试,测试结果表明该系统可以实现移动物体的检测和报警,具有一定的实用价值。本文在移动物体监控系统的设计中最大的创新点是针对移动物体检测算法的创新,将时间平均法和像素估计法的优势相融合设计研究出了一种改进像素估计法,使得移动物体的检测摆脱了虚影和噪声的干扰,检测效果达到了预想的结果,使得所设计的移动物体检测系统能实现对监控范围内移动物体图像的采集功能并且具有较高的实用性和适应性。
[Abstract]:In recent years, with the progress of society and the rapid development of science and technology, mobile object monitoring technology has also been rapidly developed.Mobile object monitoring plays a more and more important role in protecting personal safety and property safety.However, most of the traditional video surveillance is just video recording, can not extract moving objects, the follow-up work is to manually real-time viewing processing, inefficient and time-consuming.The design of mobile object monitoring system based on ARM11 is realized by the combination of mobile object detection technology and video surveillance technology.Based on the consideration of the basic function and performance of the monitoring, the sound alarm subsystem and the camera development subsystem are designed, and the design of accessing the pictures and videos of moving objects stored in the web server from the web browser is completed.The main task of software design is to propose a new mobile object detection algorithm for mobile object monitoring system.Firstly, the research status of mobile object detection, the main problems in mobile object detection algorithm and video surveillance are analyzed.After modeling and analyzing the time averaging method and the pixel estimation method, it is found that there is a virtual image in the background model obtained by the time averaging method, and the background image obtained by the pixel estimation method will misjudge the reflection of the object as the foreground.The presence of noise results in inaccurate detection results.In view of the shortcomings of the time averaging method and the pixel estimation method, this paper combines the advantages of the two algorithms, and puts forward an improved pixel estimation method, which makes the two algorithms achieve the effect of complementary advantages.The background image obtained by the improved pixel estimation method is relatively stable and can extract ideal foreground targets.It has good robustness and adaptability, and proposes an effective algorithm for moving object monitoring.Finally, the mobile object monitoring system based on ARM11 is tested. The test results show that the system can realize the detection and alarm of moving object, and it has certain practical value.In this paper, the biggest innovation in the design of mobile object monitoring system is the innovation of moving object detection algorithm. An improved pixel estimation method is developed by combining the advantages of the time averaging method and the pixel estimation method.The detection of moving objects gets rid of the interference of virtual image and noise, and the detection effect reaches the desired result.The designed mobile object detection system can realize the collection function of moving object image in the monitoring range and has higher practicability and adaptability.
【学位授予单位】：西安建筑科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP277

【参考文献】