Falling-down detection device and method

Abstract

The invention discloses a falling-down detection device and method. A gyroscope, an accelerometer and a magnetometer are respectively used for measuring to obtain angular speed data, accelerated speed data and terrestrial magnetism data; the angular speed data are utilized to solve a basic posture angle, and the accelerated speed data and the terrestrial magnetism data are utilized to carry out noise self-adapted Kalman fusion filtering on the basic posture angle to obtain a final posture angle. The angular speed data and the accelerated speed data are used tor carrying out movable and static state judgment on a user; when the user is static, whether the user lies down or not is judged according to the posture angle; if the user lies down, conscious / unconscious screening and falling-down stress assessment of lying-down movement is carried out to judge whether the user falls down or not in a combined manner; when the user falls down, an alarm is sent out. According to the falling-down detection device and method disclosed by the invention, the obtained posture angle is more accurate than that obtained by using a common method; meanwhile, the conscious / unconscious screening and the falling-down stress assessment are used for carrying out combined judgment so that a falling-down judging result is accurate and the false alarm rate is reduced.

Description

technical field [0001] The invention belongs to the technical field of fall detection, and more specifically relates to a fall detection device and method. Background technique [0002] Elderly living alone in empty nests has become a common and widely concerned social phenomenon. Due to the gradual deterioration of body functions with age and the impact of some chronic diseases, the home safety of the elderly is extremely vulnerable to the threat of sudden falls. Research on practical real-time fall detection devices has important application prospects and market value. Fall detection can rely on the optical sensor deployed in the regional environment to obtain the original image sequence, and identify changes in the state of human motion through visual methods; it can also rely on inertial devices such as accelerometers and gyroscopes deployed on the body to obtain information generated by human motion. Dynamic sensing data, fall detection through data processing and pat...

AI Technical Summary

Problems solved by technology

Using multiple sensor integration or model training methods can improve the detection quality, but it inevitably increases the complexity and power consumption, affecting real-time performance and comfort, etc.; using a single sensor has low power consumption and high real-time performance, but it often detects Insufficient quality, false positives and missed negatives occur from time to time

Images