Monitoring System For Elderly People Living Alone Based On Omni-Directional Computer Vision

With the rapid development of the society and economy, the aging problem is becoming more and more serious. The idea that the elderly are cared for by their children may be unrealistic. In recent years, monitoring the elderly using the intelligent surveillance technology has become one of the focus issues with the development of computer technology, sensor technology and communication technology.The thesis analyzes and summarizes the characteristics and shortcomings of previous monitoring systems for elderly people. A system is designed here to monitor elderly people living alone using a front-end data acquisition equipment called Omni-directional Vision Sensor (ODVS). It not only can analyze the abnormal behaviors of elderly people living alone, but also can protect the privacy of elderly people living alone. The main research contents are as follows:1. A Motion History or Energy Image algorithm (MHoEI) is used to track the object. The experimental results show that the object can be tracked steadily using the MHoEI algorithm in some cases such as objects from motion to stationary, objects from stationary to motion, objects with different poses, environment brightness changes, block, etc.2. The aspect ratio of the minimum bounding rectangle is adopted to recognize the posture. The changes of the posture and position of the elderly in the two continuous images are used together to recognize the action of the elderly. Finally, the customization environment information and position information of the elderly in the bird-view image are utilized to recognize the factual space position of elderly people.3. A Cartesian product consisting of time, posture, action and environment is utilized to describe and analyze the abnormal behaviors within the view. The routine of the elderly is used to analyze the abnormal behaviors in the blind field.4. A remote monitoring web page is designed to protect the privacy of elderly people. The data information in the database is shown in a virtual human way in the web page.In this paper, a monitoring system for elderly people living alone is designed using Java language. The implementing methods of main modules are introduced. The last part of the thesis is used to summarize the research. The existential questions are pointed out in current system. At the same time, the future work is discussed.