The design and implementation of wireless sensor networks for image surveillance

Abstract

The demand for real time monitoring is increasing rapidly. The high cost of the surveillance system makes people reluctant to adopt it in his/her places, like home or office. At present, people generally use Closed Circuit Television Video (CCTV) for monitoring. In our system, everything is upgraded and enhanced. The objective of this report is to use a latest hardware called CMUcam3 to connect to a sensor node called Tmote Sky to provide real-time photo capture function. A sensor network is formed by groups of sensor nodes. According to their small size, wireless communication, processing power and lower cost, this technology has been widely used in scientific and commercial to react with the real world. Due to its low power consumption, it can be operated by using dry cells for years. When it comes to an area where it is difficult to find an AC supply for the camera, our solution obviously helps a lot. Thanks to the broadband internet service provided by Wi-Fi hotspot and 3G, a lot of applications and services that are not possible in the past can now be provided. Users can view the images and control CMUcam3 on a sensor node through internet to and from our Java application in PC or notebook computer. Users can therefore monitor their concerned places anywhere anytime. To the best of our knowledge, this is the first implementation of a vision-enabled wireless sensor node by integrating Tmote Sky and CMUcam3.