Recognition of occupant conditions

A new recognition algorithm of occupant conditions for air conditioning systems is developed. In order to produce a comfortable thermal environment, the number and locations of occupants in a room have to be identified. We introduce our algorithm using a personal computer and a VTR. In order to get information about occupants in a room, a pyroelectric infrared ray detector to take a thermal image was developed. We introduce a segmentation algorithm of thermal images using the Fuzzy C-Means algorithm (FCM) to identify the number of occupants. We also apply fuzzy reasoning to locating occupants. According to the identified information, room temperature and wind direction can be controlled to save energy