Using fuzzy inference method to automatically detect and identify intruders from the security system

A home security system is designed to detect the illegal intruders and to warn the homeowner. When a homeowner is about to sleep or leave the house, the security system can be turned on to monitor the house and trigger the alarm if there were illegal intruders. Most security systems send alarm messages to users when the sensors are triggered, but cannot identify what the intruder is. If security systems often make false alarms when triggered by animals, people may relax their vigilance as time passes. To solve those common problems of traditional home security systems, we combine image recognition, motion detection, image processing and fuzzy inference to identify whether the illegal intruder is human, cat or dog. We use a camera to capture the images and analyze those incoming images. After obtaining the foreground object from those incoming images, we can derive its characters and then apply the fuzzy inference model to recognizing the identified object. We design the fuzzy rules to identify whether the object is human or animal. Another advantage in our model is that if we want to add more functions, we can expand the rule base. Experimental results show that the proposed fuzzy model can effectively identify the intruders.