System and method for cloud service real-time falling-down detection

Abstract

The invention discloses a system and a method for cloud service real-time falling-down detection. The system comprises an acquisition device, a mobile terminal connected with the acquisition device, a transfer device connected with the acquisition device and the mobile terminal separately, a cloud service platform connected with the transfer device, and a monitoring terminal connected with the cloud service platform, wherein the acquisition device comprises a controller, and a three-axis acceleration sensor, a three-axis gyroscope, a pressure sensor, a body temperature sensor, a heart rate sensor, a pulse sensor, a blood pressure sensor, a positioning module, a communication module and a storage module which are connected with the controller separately. According to the system and the method disclosed by the invention, the acquisition device, the mobile terminal and the transfer device form a stable data communication mode to ensure that the falling-down information of a monitored object is uploaded to the cloud service platform by virtue of the transfer device, and the real-time monitoring of quick falling-down and slow conditions of the monitored object is realized; and the data communication mode can realize functions of hierarchical processing, storage and the like of data, thereby greatly improving completeness and reliability of the data.

Description

technical field [0001] The invention belongs to the monitoring and identification of human motion behaviors and states, and in particular relates to a cloud service real-time fall detection system and method. Background technique [0002] As people grow older, people's metabolism slows down, their responses are sluggish, their physical functions decline, and they are prone to falls. According to statistics, among the elderly population over 65 years old, more than 1 / 3 of them will experience falls every year, and nearly 1 / 4 of the elderly will die within one year after the fall. According to incomplete statistics, in the accidental death of the elderly, the proportion of death caused by the fall of the elderly is as high as 2 / 3, and this proportion is as high as 70% among the elderly over 75 years old. Falls accounted for the highest proportion of deaths. Chronic falls caused by cardiovascular and other chronic diseases account for 60%, but the related research is almost z...

AI Technical Summary

Problems solved by technology

For human fall detection, it is generally divided into two types: vision-based and wearable sensor-based. Among them, the use of vision for human fall detection will be seriously affected by the external environment, such as lighting conditions, background, occlusion size, and camera quality; in addition, Due to the limited monitoring area of ​​the camera, the range of activities of the monitored elderly or patients will be limited. In the research of using wearable sensors to detect human falls, one is to use an accelerometer to detect the acceleration of human activities, and judge by setting a threshold. Falling or not, this method is difficult to distinguish between falling and people's daily strenuous activities, such as jumping, going up and down stairs, etc.

Such as CN201126620Y, CN101650869B all adopt a three-axis accelerometer to measure the acceleration of the human body, and calculate the inclination angle at the same time. The latter is to judge the angle relationship before and after a period of time when a person is impacted during the fall process to determine whether a fall has occurred. This method requires that the human body has an obvious impact during the fall process, and it is difficult to identify the elderly who suddenly fainted or fell slightly. Moreover, other The angle is calculated by acceleration. Obviously, when the human body is violently moved or disturbed by vibration, the calculated tilt angle will be seriously deviated, and the recognition rate will be seriously reduced; the other is to detect the angle of the human body through the wearable angle sensor and set the angle threshold. and time threshold to judge whether to fall, this method is difficult to distinguish normal behaviors such as bending over and lying flat

For another example, CN200941648Y and CN2909416Y use sensors to detect the inclination of the human body to judge whether a fall has occurred, and it is difficult to distinguish actions such as bending over and lying flat. In addition, due to the strong randomness and various forms of fall events, this judgment The method has a high misjudgment rate and is very unstable

Images