Occupancy Detection at Smart Home Using Real-Time Dynamic Thresholding of Flexiforce Sensor

Monitoring of the activities of the occupant is paramount important in the field of ambient-assisted living environment. Sensors are widely used to collect, store, and analyze on the continuous stream of data of the observation of their day-to-day activities. The output of any sensor may gradually change with time. It may come to a point where it becomes difficult to make a distinction between different situations from the sensor's output. Due to the change of sensor's output with time, it is hard for the system to discriminate very accurately between the regular states and the irregular states. This paper deals with the problems faced with the output of flexiforce sensors that happens when the sensor is in operation for a long period to monitor the activities of the inhabitant. A real-time dynamic thresholding has been designed and introduced to identify different situations clearly and avoid confusions regarding the output of the sensor used.