Fall Detection System For Elderly Base On Multiple Information Fusion

With the aging of the population,fall related injuries are a central problem for the elderly people,therefore many automated fall detectors have been developed.But prevalent methods are neither practical nor poor in accuracy.In this research,we design the hardware system of the old man's fall detection and proposes a new fall detection algorithm based on multi-information fusion.firstly,we analyze the characteristics of the old people's fall,then select the appropriate sensor based on the fall feature.Next,design the hardware circuit of the fall detection system.A fall detection system with a nine-axis inertial sensor is used to acquire the motion data of the pelvic position of the elderly.Finally,a fall detection algorithm is proposed,which uses a quaternion Kalman filter to fuse the information from accelerometer,gyroscope and geomagnetism so that it can form more accurate observation data.And then we use the two-stage threshold method to identify the falls,the first stage of the threshold method is posture detection to determine what posture of the old man is in.The second stage is the activity intensity analysis,mainly through the first stage to detect whether the old man is in a lying posture;if the old man is in a lying position,then the age of the elderly is measured at the time of the first 1s.Analyze whether or not you have fallen by the intensity of the movement.At the same time as the fall detection,the auxiliary amount of the heart rate sensor is also analyzed in parallel.If the old person is in a dangerous state and gives an early warning,the task of this fall detection is completed.The proposed method features low computational cost and real-time response,in addition has a nice accuracy and convenient in detect falls.