granted different types of home access based on biometric matches. As described in Scenario 1 of Section
1, some individuals may only have access to the garage or front porch while repair technicians would also
be granted access to areas of the house that need their attention. If an individual manages to enter
unauthorized areas of the house, the homeowner is notified.
Homeowners may choose to let ambient sensors run continuously and use the more intensive data-
gathering devices such as cameras only when they are out of the home. In such cases, Petersen et al. [79]
propose a method to automatically detect these situations and turn on video cameras. In this work, motion
and door sensors continuously collect data and a machine learning system is trained to map these sensor
readings onto a label indicating whether the residents are at home or away from the home. This approach
extracts features including the number of sensor firings during each five-minute interval, an indicator of
whether or not the resident is in bed, whether the door sensor was the last reading in the interval, whether
the door sensor was the first firing in the interval, and whether the last sensor in the interval emanated
from a room connected to an external door. A logistic regressor yielded a sensitivity of 0.939 and a
specificity of 0.975 on sample data collected from actual smart homes, which are strong preliminary
results supporting this approach.
While intrusion detection is a common application for security systems, much of the technology can
also be applied to health monitoring and assistance as well. In the case of work by Dodge et al. [80], by
Hodges et al. [81], by Dawadi et al. [82]
, and by Lotfi et al. [83], unexpected behavioral patterns are
viewed as a health risk for individuals who are at risk of cognitive decline. These researchers have found
that an increase in the number of activity anomalies and variation in behavior patterns such as activity
times and walking speed are correlated with changes in cognitive health. As in the case with the intrusion
detection research, these findings provide insights that can be used by smart homes in order to keep
residents safe. For example, residents and their caregivers can use this information to change the level of
care that the individual needs.
In research by Ali et al. [84] and of Das et al. [85], threats are detected in the form of abnormalities in
how residents perform their daily activities. For many individuals, these variations would not be
considered a risk. However, for individuals with memory limitations, performing daily activities
independently is critical. Functional impairment has been associated with increased health care use and
Figure 2. Technologies found in a secure smart home.
剩余22页未读，继续阅读
无敌篓子
- 粉丝: 1
- 资源: 4
我的内容管理
展开
最新资源
- ASP.NET数据库高级操作：SQLHelper与数据源控件
- Windows98/2000驱动程序开发指南
- FreeMarker入门到精通教程
- 1800mm冷轧机板形控制性能仿真分析
- 经验模式分解：非平稳信号处理的新突破
- Spring框架3.0官方参考文档：依赖注入与核心模块解析
- 电阻器与电位器详解：类型、命名与应用
- Office技巧大揭秘：Word、Excel、PPT高效操作
- TCS3200D: 可编程色彩光频转换器解析
- 基于TCS230的精准便携式调色仪系统设计详解
- WiMAX与LTE：谁将引领移动宽带互联网?
- SAS-2.1规范草案：串行连接SCSI技术标准
- C#编程学习：手机电子书TXT版
- SQL全效操作指南：数据、控制与程序化
- 单片机复位电路设计与电源干扰处理
- CS5460A单相功率电能芯片：原理、应用与精度分析
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
