A fall detection method based on wearable devices

Abstract

The present invention discloses a fall detection method based on a wearable device. The wearable device includes: a motion sensor and a heart rate measurement sensor; the detection method includes: capturing a suspected fall segment signal through the motion sensor; according to the suspected fall segment signal, turning on the heart rate measurement sensor; detecting whether some or all of the indicators in the post-fall heart rate indicator group HR2 meet a preset condition through the heart rate measurement sensor; the preset condition is: exceeding a first preset range of a preset heart rate indicator group HR1 or being lower than a second preset range; if the preset condition is met, a fall warning is issued for the user wearing the wearable device. This method solves the problem that fall detection is prone to false judgment. By using multi-sensor fusion for fall detection, it can reduce the false alarm rate, can reduce the disturbance to the user, and can also avoid false alarms caused by dropping the wearable device.

Description

Technical Field

[0001] The present invention relates to the technical fields of wearable devices and fall detection, and particularly relates to a fall detection method based on a wearable device. Background Art

[0002] In recent years, various technologies combined with machine learning and deep learning algorithms have been widely applied to the field of fall detection. Almost all implementation steps of fall detection systems are: perception, data processing, fall event recognition, and emergency alarm. Currently, there are mainly three common types of fall detection systems: wearable device-based, environment sensor-based, and video sensor-based human fall detection systems. Among them, the wearable device market is continuously expanding, and there are a large number of elderly users. Wearable devices provide more convenience and protection for elderly users.

[0003] However, the current fall detection of wearable devices has a high false alarm rate, and users need to confirm and cancel false alarms, which brings certain inconvenience to users' lives. Moreover, the existing fall detection technologies also have the following disadvantages:

[0004] ① The false alarm rate of fall determination based only on motion sensors is relatively high;

[0005] ② It is difficult for a barometric pressure sensor to detect changes within 1 m;

[0006] ③ It brings more inconvenience to users to confirm whether they have fallen;

[0007] ④ It can only target relatively severe falls;

[0008] Therefore, for the problems of high false alarm rate, low detection accuracy, inconvenient operation, and inability to accurately detect some falls in single-sensor determination, it is urgent for practitioners in the same field to solve. Summary of the Invention

[0009] The purpose of the present invention is to provide a fall detection method based on a wearable device that at least partially solves the above technical problems. This method has a relatively low false alarm rate for fall determination and can reduce the interference with users' normal lives.

[0010] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0011] In a first aspect, the present invention provides a fall detection method based on a wearable device. The wearable device includes: a motion sensor and a heart rate measurement sensor; the detection method includes:

[0012] Capturing a suspected fall segment signal through the motion sensor;

[0013] Starting the heart rate measurement sensor according to the suspected fall segment signal;

[0014] Detect whether some or all of the indicators in the heart rate indicator group HR2 after a fall meet the preset conditions through the heart rate measurement sensor; the preset conditions are: exceeding the first preset range of the preset heart rate indicator group HR1 or being lower than the second preset range;

[0015] If the preset conditions are met, issue a fall warning for the user wearing the wearable device.

[0016] Further, the preset heart rate indicator group HR1 is the heart rate indicator group measured by the heart rate measurement sensor under the daily activity intensity of the user: or the default heart rate indicator group set according to the user's physical condition.

[0017] Further, after the preset conditions are met, the detection method further includes:

[0018] Determine the suspected fall moment according to the suspected fall segment signal, intercept several segment signals including the suspected fall moment, and determine whether a fall exists;

[0019] When it is determined that a fall exists, execute the step of issuing a fall warning for the user wearing the wearable device.

[0020] Further, determining the suspected fall moment according to the suspected fall segment signal, intercepting several segment signals including the suspected fall moment, and determining whether a fall exists; includes:

[0021] Obtain the suspected fall segment signal a1 with a duration of L1 according to the motion sensor, and cache the suspected fall segment signal a0 with a time L2 longer than the duration L1; where L1 = t2 - t1; t1 represents the starting moment, and t2 represents the ending moment;

[0022] Determine the specific moment T - trip1 of the suspected fall according to the suspected fall segment signal a1;

[0023] Obtain several time sliding window segment signals A - addN before and after the specific moment T - trip1 in the suspected fall segment signal a0; several of the time sliding window segment signals A - addN all include the specific moment T - trip1 and do not overlap with the suspected fall segment signal a1;

[0024] Judge whether a fall exists in several of the time sliding window segment signals A - addN.

[0025] Further, after the preset conditions are met, the detection method further includes:

[0026] Determine the suspected fall moment according to the suspected fall segment signal, intercept the front and back two segment signals that do not include the suspected fall moment, and determine whether the motion state has changed;

[0027] When the determination changes, perform the step of issuing a fall warning for the user wearing the wearable device.

[0028] In a second aspect, an embodiment of the present invention further provides a fall detection method based on a wearable device. The wearable device includes: a motion sensor; and the detection method includes:

[0029] Capture a suspected fall segment signal through the motion sensor;

[0030] Determine a suspected fall moment according to the suspected fall segment signal, intercept a plurality of segment signals including the suspected fall moment, and determine whether a fall exists;

[0031] When it is determined that a fall exists, issue a fall warning for the user wearing the wearable device.

[0032] Further, determining a suspected fall moment according to the suspected fall segment signal, intercepting a plurality of segment signals including the suspected fall moment, and determining whether a fall exists; includes:

[0033] Obtain a suspected fall segment signal a1 with a duration L1 from the motion sensor, and cache a suspected fall segment signal a0 with a time L2 longer than the duration L1; where L1 = t2 - t1; t1 represents the start moment, and t2 represents the end moment;

[0034] Determine the specific moment T - trip1 of the suspected fall according to the suspected fall segment signal a1;

[0035] Obtain a plurality of time sliding window segment signals A - addN before and after the specific moment T - trip1 in the suspected fall segment signal a0; a plurality of the time sliding window segment signals A - addN all include the specific moment T - trip1 and do not overlap with the suspected fall segment signal a1;

[0036] Judge whether a fall exists in a plurality of the time sliding window segment signals A - addN.

[0037] Further, after the step of when it is determined that a fall exists, the detection method further includes:

[0038] Intercept the front and back two segment signals that do not include the suspected fall moment according to the suspected fall moment, and determine whether the motion state has changed;

[0039] When the determination changes, perform the step of issuing a fall warning for the user wearing the wearable device.

[0040] In a third aspect, an embodiment of the present invention further provides a fall detection method based on a wearable device. The wearable device includes: a motion sensor; and the detection method includes:

[0041] Capture a suspected fall segment signal through the motion sensor;

[0042] Determine the suspected fall moment according to the suspected fall segment signal, intercept the front and back two segment signals that do not include the suspected fall moment, and determine whether the motion state has changed;

[0043] When it is determined that a change has occurred, issue a fall warning for the user wearing the wearable device.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] A fall detection method based on a wearable device, the wearable device includes: a motion sensor and a heart rate measurement sensor; the detection method includes: capturing a suspected fall segment signal through the motion sensor; according to the suspected fall segment signal, turning on the heart rate measurement sensor; detecting whether some or all of the indicators in the post-fall heart rate indicator group HR2 meet a preset condition through the heart rate measurement sensor; the preset condition is: exceeding the first preset range of the preset heart rate indicator group HR1 or being lower than the second preset range; if the preset condition is met, issue a fall warning for the user wearing the wearable device. This method has a low false alarm rate for fall determination and can reduce the interference with the normal life of the user.

[0046] Furthermore, after the preset condition is met, it is also possible to further combine and intercept several segment signals including the fall moment to determine whether a fall also exists; it is possible to avoid false alarms caused by the wearable device falling and reduce the false alarm rate.

[0047] Even further, it is also possible to combine and intercept the front and back two segment signals that do not include the fall moment to determine whether the motion states are different; while improving the detection accuracy, it can also reduce the interference with the user. Brief Description of the Drawings

[0048] Figure 1 It is a flowchart of the fall detection method based on a wearable device in Embodiment 1;

[0049] Figure 2 It is an overall flowchart of the fall detection method based on a wearable device;

[0050] Figure 3 It is a schematic diagram of taking data of the motion signal collected by the motion sensor;

[0051] Figure 4 It is a schematic diagram of taking data of several time sliding window segment signals A-addN;

[0052] Figure 5 It is a schematic diagram of taking data of two time sliding window segment signals A-after and A-before;

[0053] Figure 6 Flow chart of the fall detection method based on a wearable device according to Embodiment 2;

[0054] Figure 7 Flow chart of the fall detection method based on a wearable device according to Embodiment 3. Detailed implementation manners

[0055] To make the technical means, creative features, achieved purposes and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific implementation manners.

[0056] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "upper", "lower", "inner", "outer", "front end", "rear end", "both ends", "one end", "the other end", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0057] In the description of the present invention, it should be noted that unless otherwise clearly specified and limited, the terms "installed", "provided with", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0058] Embodiment 1:

[0059] A fall detection method based on a wearable device provided by the present invention. The wearable device can be placed at different positions on the body, usually worn on the wrist, and includes one or more groups of motion sensors and a heart rate measurement sensor that can measure the heart rate of a person;

[0060] Referring to Figure 1 as shown, the detection method includes:

[0061] S11. Capturing a suspected fall segment signal through the motion sensor;

[0062] S12. Turning on the heart rate measurement sensor according to the suspected fall segment signal;

[0063] S13, detecting, by the heart rate measurement sensor, whether all or part of the indicators in the heart rate indicator group HR2 after the fall meet a preset condition; the preset condition is: exceeding a first preset range of the preset heart rate indicator group HR1 or being lower than a second preset range;

[0064] S14: If the preset conditions are met, a fall warning is issued to the user wearing the wearable device.

[0065] Heart rate refers to the frequency of heart beats in one minute. The standard heart rate range for adults at rest is 60-100 beats / minute. Among them, the heart rate range of the elderly is also 60-100 beats / minute. The heart rate of the elderly varies from person to person, and the heart rate is adapted to the internal environment of the human body. The body functions of the elderly have declined to a certain extent, the basal metabolic rate is low, and the amount of exercise is relatively small. Therefore, under normal circumstances, the heart rate is relatively slow, mostly around 60-80 beats / minute. In addition, there are also many elderly people whose heart rate may be around 50-60 beats / minute. If the patient does not have any discomfort symptoms, it can also be regarded as a normal heart rate.

[0066] In this embodiment, the heart rate index group HR1 under the intensity of daily activities is measured based on the wrist-worn heart rate measurement sensor. If it is not measured, it will default to several initial values, which can also be set by the user according to his own conditions; this is used as the basis for determining whether the heart rate is abnormal after a fall.

[0067] The motion sensor is used to capture suspected fall clips. When a suspected fall clip occurs, the heart rate measurement sensor is turned on to detect whether some or all of the indicators in the heart rate indicator group HR2 after the fall exceed the user's normal heart rate indicator group HR1, a certain threshold group RANGE1, or are lower than the threshold group RANGE2. For example, the heart rate indicator group HR1 of elderly A is 60-80 beats / minute. If the heart rate exceeds the threshold range, when it is lower than 60 beats / minute and is within the range of 40-50 beats / minute; or when it exceeds 80 beats / minute and is within the range of 100-120 beats / minute, it is determined that elderly A has fallen, and an early warning of the user's fall is issued so that the user can get timely assistance.

[0068] In order to further improve the accuracy of detection, after the above preset conditions are met, one or two judgment conditions can be added, and the two judgment conditions are not in order. The overall process of fall detection is as follows: Figure 2 As shown:

[0069] First: determine the suspected fall moment according to the suspected fall segment signal, intercept several segment signals including the suspected fall moment, and judge whether there is a fall; when it is judged that there is a fall, execute the step of issuing a fall warning to the user wearing the wearable device.

[0070] Specifically, they include:

[0071] S101. As shown in Figure 3 , obtain a suspected fall segment signal a1 with a duration of L1 from the motion sensor, and cache a suspected fall segment signal a0 with a time duration L2 that is longer than L1; where L1 = t2 - t1; t1 represents the starting moment, and t2 represents the ending moment.

[0072] S102. Determine the specific moment T - trip1 of the suspected fall according to the suspected fall segment signal a1, as shown in Figure 3 .

[0073] S103. As shown in Figure 4 , obtain several time sliding window segment signals A - addN (A - add1, A - add2, A - add3...) before and after the specific moment T - trip1 from the suspected fall segment signal a0; several time sliding window segment signals A - addN all contain the specific moment T - trip1 and do not overlap with the suspected fall segment signal a1.

[0074] S104. Determine whether there is a fall in several time sliding window segment signals A - addN. When it is determined that there is a fall, execute the step of issuing a fall warning for the user wearing the wearable device.

[0075] Based on the wrist - worn device, obtain a suspected fall segment signal a1 with a certain duration L1 (L1 = t2 - t1) from the motion sensor (such as the dotted part in Figure 3 ), and cache an acceleration signal a0 with a longer time L2 (such as the dash - dotted part in Figure 3 ). Based on a certain moment sliding window signal a1 (suspected fall segment signal), determine whether there is a fall in the current state and determine the specific moment T - trip1 of the fall. The specific value - taking method is as shown in Figure 3 . If it is determined that there is a fall, at the tc moment after a certain moment (such as tc - t2) after t2, take several L1 - length signals A - addN (A - add1, A - add2, A - add3) that contain the fall moment T - trip1 and do not overlap with the a1 signal. When there is a fall in these several time sliding window segment signals A - addN, execute the step of issuing a fall warning for the user wearing the wearable device.

[0076] Second: As shown in Figure 5 , determine the suspected fall moment according to the suspected fall segment signal, intercept two segment signals A - befor and A - after that do not contain the suspected fall moment T - trip1, and determine whether the motion state has changed; when it is determined that the change has occurred, execute the step of issuing a fall warning for the user wearing the wearable device.

[0077] As shown in Figure 2As shown, the overall process of fall detection includes three conditions: heart rate judgment of fall, sliding window judgment including the moment of fall, and excluding the fragments before and after the moment of fall. The latter two conditions can be executed in any order. In the case of insufficient computing power, the two judgment conditions of including the sliding window judgment at the moment of fall and excluding the fragments before and after the moment of fall can be selectively removed.

[0078] An embodiment of the present invention provides a fall detection method based on a wearable device, wherein the wearable device can be placed at different positions on the body, usually worn on the wrist, and includes one or more groups of motion sensors and sensors that can measure a person's heart rate;

[0079] The detection process includes:

[0080] 1) The motion sensor detects the fall signal characteristics, which are higher than the specified value, and turns on the heart rate measurement sensor for measurement;

[0081] 2) Part or all of the differences between the heart rate index group measured by the heart rate measurement sensor and the heart rate index group during normal wearing are lower than or higher than the specified threshold group;

[0082] 3) The content that can be added includes:

[0083] a) The movement state is different before and after the fall;

[0084] b) intercepting several clips containing the moment of falling and also determining that there is a fall;

[0085] 4) The content in (3) can be flexibly configured according to computing power.

[0086] This method solves the problem of misjudgment in fall detection. It uses multi-sensor fusion for fall detection, which can reduce the false alarm rate, reduce disturbance to users, and avoid false alarms of falls caused by dropping wearable devices.

[0087] Embodiment 2:

[0088] The present invention also provides a fall detection method based on a wearable device, and the wearable device has the same configuration as that of Example 1.

[0089] Reference Figure 6 As shown, the detection method includes:

[0090] S21, capturing a suspected fall segment signal by the motion sensor;

[0091] S22, determining the suspected fall moment according to the suspected fall segment signal, intercepting several segment signals including the suspected fall moment, and judging whether there is a fall;

[0092] S23. When it is determined that a fall has occurred, a fall warning is issued for the user wearing the wearable device.

[0093] In this embodiment, the suspected fall moment is determined according to the suspected fall segment signal, and several segment signals including the suspected fall moment are intercepted to determine whether a fall has occurred; when it is determined that a fall has occurred, the step of issuing a fall warning for the user wearing the wearable device is executed.

[0094] Specifically, the above step S22 includes:

[0095] S201. As Figure 3 shown, a suspected fall segment signal a1 with a duration L1 is obtained according to the motion sensor, and a suspected fall segment signal a0 with a longer time L2 than the duration L1 is cached; where L1 = t2 - t1; t1 represents the starting moment, and t2 represents the ending moment;

[0096] S202. According to the suspected fall segment signal a1, the specific moment T - trip1 of the suspected fall is determined, as Figure 3 shown;

[0097] S203. As Figure 4 shown, several time sliding window segment signals A - addN (A - add1, A - add2, A - add3...) before and after the specific moment T - trip1 are obtained from the suspected fall segment signal a0; several time sliding window segment signals A - addN all include the specific moment T - trip1 and do not overlap with the suspected fall segment signal a1;

[0098] S204. Determine whether a fall exists in several time sliding window segment signals A - addN. When it is determined that a fall has occurred, the step of issuing a fall warning for the user wearing the wearable device is executed.

[0099] Based on the wrist - worn device, a suspected fall segment signal a1 with a certain duration L1 (L1 = t2 - t1) is obtained by the motion sensor (such as the dotted part in Figure 3 ), and the acceleration signal a0 with a longer time L2 is cached (such as the dash - dotted part in Figure 3 ). Based on a certain moment sliding window signal a1 (suspected fall segment signal), it is determined whether a fall has occurred in the current state and the specific moment T - trip1 of the fall is judged. The specific value - taking method is as Figure 3If a fall is determined, then at time tc after a certain time after t2 (such as tc-t2), several L1 length signals A-addN (A-add1, A-add2, A-add3) that include the fall time T-trip1 and do not overlap with the a1 signal are taken. When these several time sliding window segment signals A-addN also indicate a fall, the step of issuing a fall warning to the user wearing the wearable device is executed.

[0100] Similarly, in this embodiment, in order to further improve the accuracy of detection, the wearable device includes: a motion sensor and a heart rate measurement sensor; based on the condition: including several fragment signals of the suspected fall moment to determine whether there is a fall, two more conditions can be added, such as the heart rate judgment of the fall in Example 1, and the judgment of the fall by not including the fragments before and after the fall moment. The latter two conditions can be executed in no particular order. For the case of insufficient computing power, the two judgment conditions or any one of the judgment conditions can be selectively removed; the specific implementation process can refer to the content of the above-mentioned Example 1, and the repeated parts will not be repeated.

[0101] Embodiment 3:

[0102] The present invention further provides a fall detection method based on a wearable device, and the wearable device has the same configuration as that of Example 1.

[0103] Reference Figure 7 As shown, the detection method includes:

[0104] S31, capturing a suspected fall segment signal by the motion sensor;

[0105] S32, determining the suspected fall moment according to the suspected fall segment signal, intercepting the two segment signals before and after that do not include the suspected fall moment, and determining whether the motion state has changed;

[0106] S33: When the determination is changed, a fall warning is issued to the user wearing the wearable device.

[0107] In this embodiment, Figure 5 As shown, the suspected fall moment is determined according to the suspected fall segment signal, and the two segment signals A-befor and A-after that do not include the suspected fall moment T-trip1 are intercepted to determine whether the motion state has changed; when it is determined that a change has occurred, the step of issuing a fall warning to the user wearing the wearable device is executed.

[0108] Similarly, in this embodiment, in order to further improve the accuracy of detection, the wearable device includes: a motion sensor and a heart rate measurement sensor; on the premise of determining whether the motion state has changed based on the condition of not including the segments before and after the moment of falling, two conditions such as heart rate determination of falling and sliding window determination including the moment of falling in Embodiment 1 can be added. Among them, the latter two conditions can be executed in any order. For the case of insufficient computing power, the two judgment conditions or any one of the judgment conditions can be selectively removed. The specific implementation process can refer to the content of Embodiment 1 above, and the repeated parts will not be elaborated here.

[0109] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. A fall detection method based on a wearable device, characterized in that, the wearable device includes: a motion sensor and a heart rate measurement sensor; the detection method includes: capturing a suspected fall segment signal through the motion sensor; activating the heart rate measurement sensor according to the suspected fall segment signal; detecting whether some or all of the indicators in the post-fall heart rate indicator group HR2 meet a preset condition through the heart rate measurement sensor; the preset condition is: exceeding a first preset range of a preset heart rate indicator group HR1 or being lower than a second preset range; if the preset condition is met, determining a suspected fall moment according to the suspected fall segment signal, intercepting a plurality of segment signals including the suspected fall moment, and determining whether there is a fall in all of them; when it is determined that there is a fall in all of them, performing the step of issuing a fall warning for the user wearing the wearable device; wherein, if the preset condition is met, determining a suspected fall moment according to the suspected fall segment signal, intercepting a plurality of segment signals including the suspected fall moment, and determining whether there is a fall in all of them, includes: obtaining a suspected fall segment signal a1 with a duration L1 through the motion sensor, and caching a suspected fall segment signal a0 with a time L2 longer than the duration L1; where L1 = t2 - t1; t1 represents the start moment, and t2 represents the end moment; determining a specific moment T-trip1 of the suspected fall according to the suspected fall segment signal a1; acquiring a plurality of time sliding window segment signals A-addN before and after the specific moment T-trip1 in the suspected fall segment signal a0; a plurality of the time sliding window segment signals A-addN all include the specific moment T-trip1 and do not overlap with the suspected fall segment signal a1; determining whether there is a fall in all of the plurality of time sliding window segment signals A-addN.

2. The fall detection method based on a wearable device according to claim 1, characterized in that, the preset heart rate indicator group HR1 is the heart rate indicator group measured by the heart rate measurement sensor under the daily activity intensity of the user: or the default heart rate indicator group set according to the user's physical condition.

3. The fall detection method based on a wearable device according to claim 1, characterized in that, after the step of when it is determined that there is a fall in all of them, the detection method further includes: determining a suspected fall moment according to the suspected fall segment signal, intercepting two segment signals before and after that do not include the suspected fall moment, and determining whether the motion state has changed; when it is determined that the change has occurred, performing the step of issuing a fall warning for the user wearing the wearable device.

4. A fall detection method based on a wearable device, characterized in that, the wearable device includes: a motion sensor; the detection method includes: capturing a suspected fall segment signal through the motion sensor; determining a suspected fall moment according to the suspected fall segment signal, intercepting a plurality of segment signals including the suspected fall moment, and determining whether there is a fall in all of them; when it is determined that there is a fall in all of them, then issuing a fall warning for the user wearing the wearable device; Among them, a suspected fall moment is determined according to the suspected fall segment signal, and several segment signals including the suspected fall moment are intercepted to determine whether a fall exists in all of them, including: A suspected fall segment signal a1 with a duration L1 is obtained according to the motion sensor, and a suspected fall segment signal a0 with a time L2 longer than the duration L1 is cached; where L1 = t2 - t1; t1 represents the starting moment and t2 represents the ending moment; According to the suspected fall segment signal a1, a specific moment T-trip1 of the suspected fall is determined; A plurality of time sliding window segment signals A-addN before and after the specific moment T-trip1 are obtained from the suspected fall segment signal a0; all of the plurality of time sliding window segment signals A-addN include the specific moment T-trip1 and do not overlap with the suspected fall segment signal a1; Judge whether a fall exists in all of the plurality of time sliding window segment signals A-addN.

5. A fall detection method based on a wearable device according to claim 4, characterized in that, After the step of judging that a fall exists in all of them, the detection method further includes: According to the suspected fall moment, two segment signals before and after that do not include the suspected fall moment are intercepted to determine whether the motion state has changed; When it is determined that a change has occurred, the step of issuing a fall warning for the user wearing the wearable device is executed.

6. A fall detection method based on a wearable device, characterized in that, The wearable device includes: a motion sensor; the detection method includes: Capturing a suspected fall segment signal through the motion sensor; Determining a suspected fall moment according to the suspected fall segment signal, intercepting two segment signals before and after that do not include the suspected fall moment, and determining whether the motion state has changed; When it is determined that a change has occurred, then determining a suspected fall moment according to the suspected fall segment signal, intercepting several segment signals including the suspected fall moment, and judging whether a fall exists in all of them; specifically including: A suspected fall segment signal a1 with a duration L1 is obtained according to the motion sensor, and a suspected fall segment signal a0 with a time L2 longer than the duration L1 is cached; where L1 = t2 - t1; t1 represents the starting moment and t2 represents the ending moment; According to the suspected fall segment signal a1, a specific moment T-trip1 of the suspected fall is determined; A plurality of time sliding window segment signals A-addN before and after the specific moment T-trip1 are obtained from the suspected fall segment signal a0; all of the plurality of time sliding window segment signals A-addN include the specific moment T-trip1 and do not overlap with the suspected fall segment signal a1; Judge whether a fall exists in all of the plurality of time sliding window segment signals A-addN; When it is judged that a fall exists in all of them, a fall warning is issued for the user wearing the wearable device.

Images