Fall detection system based on kinect sensor

Abstract

The invention provides a fall detection system based on a kinect sensor. The system comprises a kinect sensor, a human skeleton recognition module, a camera shooting operation visual monitoring moduleand a LabVIEW Real-Time tool. The Kinect sensor is used as a hardware basis; extracting angle features and distance features of human body behavior features as feature vectors of combined features; aKNN algorithm is used as a theoretical basis; visual monitoring information is obtained through multiple cameras. fall posture detection is achieved based on a Labview virtual software platform; meanwhile, real-time communication with a mobile phone is realized through a wireless communication technology; the old people fall down for 24 hours and are monitored in real time, automatic alarming isconducted on old people users with abnormal fall-down indexes, family members or nursing personnel are reminded to rescue the old people in time, and meanwhile storage of visual information can assistmedical workers in quick symptomatic treatment of old patients.

Description

technical field [0001] The invention relates to the technical field of medical detection, in particular to a fall detection system based on a kinect sensor. Background technique [0002] In daily life, the rapid growth of the elderly population has brought a series of problems. As the age increases, the central nervous system, sensory function, bones, joints, ligaments, muscles and other functions of the elderly will degenerate. In addition, vision, intelligence, and reaction ability will also decline, and stability and balance functions will be weakened. Falls in the elderly can lead to serious consequences such as fractures, craniocerebral injury, and even death, which seriously threatens the physical and mental health, daily activities and independent living ability of the elderly, and also increases the burden on the family and society. Especially for the elderly living alone, they bear more risks of falls. Falls not only easily cause physical disability and organ dama...

AI Technical Summary

Problems solved by technology

However, a single sensor is often unable to support complex body movement recognition, and wearing too many sensors will greatly reduce the wearing comfort, increase equipment costs and maintenance costs

Environmental fall detection does not need to be worn, and it is not easy to expose privacy, but it is greatly affected by the indoor environment (noise, etc.), and the false alarm rate is high

However, the main disadvantage of wearable devices is that they need to be worn on the human body at all times in daily use, which brings constraints to the lives of the elderly.

Among them, it is difficult to achieve the purpose of fall detection if you cannot wear it when you are bathing, resting, washing, or getting up at night, or you are unwilling to wear it because it is troublesome to wear. The traditional wearable fall recognition technology will cause human discomfort. The reporting rate is very high; the video surveillance system of the monocular vision sensor, due to the occlusion of objects in the home environment, will have the problem of missing detection

Images