Human Situation Monitoring System

The need for personal safety and security is rising everywhere: on the street, at home, in school. If we only had a “watchdog” – one that would be everywhere and see everything, one that would raise the appropriate alarm in real time – yet a watchdog that is unobtrusive, doesn’t compromise people’s privacy, doesn’t generate false alarms, and doesn’t cost a lot. Can we dream of such a watchdog? In the HuSIMS project we aim to show that this is indeed feasible. From year to year, our sense of security is diminished: in the city center, at the office, during travel, on vacation, in our homes. Violence, crime, accidents and the fear of terrorism are everywhere. How can we improve this situation? Doubling the police force would not be welcomed. Placing millions of video cameras intrudes on personal privacy, and has proved to be ineffective [California State Library report CRB-97-005June 1997, California Research Bureau]. Access limitations, metal detectors and electronic checkpoints cannot be tolerated except under extreme conditions such as airports. HuSIMS will develop a system that can improve safety and security as well as increase people’s sense of security. HuSIMS will achieve this by improving the effectiveness and timeliness of response to irregular situations: ambulances to accident victims, policemen to control a scene of violence, and civil professionals to solve infrastructure crisis (electricity, gas, water etc). Great attention for respecting personal privacy is one of our main considerations. Finally, the system must be effective and affordable, reusing the existing infrastructure whereas possible, such as in hospitals and hotels. How do we plan to achieve all that? Our technical solution is built of several novel components, together with available components, all integrated in a new way and invested with intelligence at various layers. The main components of HuSIMS are:

  • A dense network of low-cost, low-resolution, intelligent visual sensors, “a sensor on every lamppost”. The sensors interpret their local area and send scene indication data packets (not video) to the backend, over a low bandwidth network.
  • A seamless outdoor wireless network connecting the sensors to the backend system.
  • A complementary wired network inside the houses and buildings to reinforce the connection among the sensors and the backend system as well as with the in-home broadband network.
  • A scene interpretation system running on the backend, optimized per each sensor.
  • An decision making system analyzing the data from all the sensors and interpreting the situation.
  •  An interface to responders (control room, mobile devices) providing interpreted alerts with verification video clips.

How will this system perform? Let’s take a city quarter as an example. 200-300 visual sensors are deployed around the streets and alleys of the quarter. They are small and unobtrusive, and communicate wirelessly with the backend control center located somewhere in the city. Each sensor monitors a region of about 30X30 m with low resolution (320X240 pixels). The low resolution usually does not allow recognizing faces or people. The sensors measure general, rather then specific, parameters of the scene: level and direction of motion of people/cars, distribution of people / vehicles, and clustering thereof. The sensor will then send data packets with the information extracted from the video. These packets require only little bandwidth, but contain sufficient information for the scene analysis engine at the backend to enable it to make an informed decision on whether that street is quiet and normal, or there’s something there that merits a second look from a controller in the control room. In contrast with outdoor places, a HuSIMS installation in a hospital or large hotel will be interconnected via zigbee to the wired network or directly to the backend control center. In Home surveillance, this will allow citizens to monitor their homes or small offices in real time remotely, or enable security companies to do it in their control rooms. The sensor technology will be built around Emza’s recognized WiseEye ® visual sensor and its underlying Reduced Recognition Architecture (RRA). Its inherent features of optimal sensitivity even in the face of extreme visual conditions (moving clouds and shadows, rain, snow etc), coupled with low cost, simplicity of installation and low power requirement, make it an ideal platform for developing the intelligent sensors required by HuSIMS. Upon an alert, the controller / end user can request real-time video from one or more sensors to improve his understanding of the situation and enable making a better decision. The system will continually train and improve its performance using it’s built in learning and optimization system The system will be cost effective: the sensors cost less than standard video cameras; their installation requires very little infrastructure (only hooking up to 12VDC); and the operation allows one or two controllers to efficiently monitor a very large area. If we compare this to the cost of adding a few policemen, the attractiveness of this solution becomes evident. For in-house/in-building surveillance sensor cost can drop dramatically as they will not require outdoors weather resistance, and will be able to reuse the existing wiring. As the world becomes less secure, it is looking for effective, affordable security solutions. The HuSIMS platform offers an effective, affordable solution that is unparalleled.