Method to determine a pressure noise metric for fall detection systems
A wearable fall-detection device has a variety of sensors, including a pressure sensor, that provide signals for sampling environmental conditions acting on the device. An average of pressure data samples is used to determine a resultant that may indicate an amount of noise in a pressure data signal, and statistical analysis of the noise and the pressure signal average may be used to determine a confidence estimate value that indicates a level of confidence in the amount of noise that a pressure signal is subject to, or includes. The confidence estimate and known fall data, such as change in pressure between a person standing and lying, can create a threshold function that may adapt according to sampled data thus providing a customizable (either statically or dynamically) threshold function for comparing sensor data against rather than comparing data with just a linear threshold function.
1. A method, comprising:
storing, by a device for detecting a fall event, barometric pressure data samples in a first buffer,
the device comprising an accelerometer, and
the barometric pressure data samples being barometric pressure data sampled from a signal generated by a barometric pressure sensor;
storing, by the device and in a second buffer, acceleration data samples,
the acceleration data samples being acceleration data sampled from a signal generated by the accelerometer,
each acceleration data sample corresponding to a respective barometric pressure data sample, and
each acceleration data sample being substantially temporally sampled with the respective barometric pressure data sample;
determining, by the device, that a potential fall event has occurred based on a magnitude of a particular acceleration data sample exceeding a predetermined acceleration magnitude criterion;
selecting, by the device, first barometric pressure data samples in the first buffer having a time of sampling within a predetermined period of time of a sampling of the particular acceleration data sample;
evaluating, by the device, the first barometric pressure data samples to determine confidence estimate values associated with the first barometric pressure data samples;
outputting, by the device, the confidence estimate values;
calculating, by the device, a barometric pressure threshold based on the confidence estimate values;
selecting, by the device, second barometric pressure data samples having a time of sampling during a period of time after the sampling of the particular acceleration data sample;
determining, by the device and based on selecting the second barometric pressure data samples, a change in barometric pressure that occurred between a time before the potential fall event and a time after the potential fall event;
comparing, by the device, the change in barometric pressure and the barometric pressure threshold to determine whether the potential fall event is the fall event; and
outputting, by the device, a signal indicating that the potential fall event is the fall event based on the change in barometric pressure satisfying the barometric pressure threshold.
2. The method of claim 1 , wherein the acceleration data samples are further generated by at least one of:
a gyroscope,
a manometer, or
a proximity sensor.
3. The method of claim 1 , wherein a confidence estimate value, of the confidence estimate values, is a standard deviation of the first barometric pressure data samples in the first buffer.
4. The method of claim 1 , wherein determining that the potential fall event has occurred based on the magnitude of the particular acceleration data sample exceeding the predetermined acceleration magnitude criterion includes at least one of:
determining changes in orientation of the device,
determining a free-fall based on acceleration magnitude signals,
determining an indication in accelerometer data that indicate an impact with a ground,
determining a duration of the indication,
determining a duration of a period of free-fall,
determining a standard deviation of a portion of information associated with one or more of an X-axis, a Y-axis, a Z-axis, or magnitude buffers,
determining a mean of the portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determining an area of a spike portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determining an area of a free-fall portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determining a rate of change of the spike portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determining a rate of change of the free-fall portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determining a mode of data in the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determining a median of data in the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determining a minimum of data in one or more X-axis, Y-axis, Z-axis, or magnitude threshold values being met, or
determining a peak X-axis, Y-axis, Z-axis, or magnitude threshold value being met.
5. The method of claim 1 , wherein a confidence estimate value corresponds to an amount of noise included in the signal generated by the barometric pressure sensor.
6. The method of claim 1 , wherein the change in barometric pressure corresponds to a change in altitude of the device.
7. A device, comprising:
a processor to:
store, for detecting a fall event, barometric pressure data samples in a first buffer,
the device comprising an accelerometer, and
the barometric pressure data samples being barometric pressure data sampled from a signal generated by a barometric pressure sensor;
store, in a second buffer, acceleration data samples,
the acceleration data samples being acceleration data sampled from a signal generated by the accelerometer,
each acceleration data sample corresponding to a respective barometric pressure data sample, and
each acceleration data sample being substantially temporally sampled with the respective barometric pressure data sample;
determine that a potential fall event has occurred based on a magnitude of a particular acceleration data sample exceeding a predetermined acceleration magnitude criterion;
select first barometric pressure data samples in the first buffer,
the first barometric pressure data samples having a time of sampling within a predetermined period of time of a sampling of the particular acceleration data sample;
evaluate the first barometric pressure data samples to determine confidence estimate values associated with the first barometric pressure data samples;
output the confidence estimate values;
calculate a barometric pressure threshold based on the confidence estimate values;
select second barometric pressure data samples having a time of sampling during a period of time after the sampling of the particular acceleration data sample;
determine, based on selecting the second barometric pressure data samples, a change in barometric pressure that occurred between a time before the potential fall event and a time after the potential fall event;
compare the change in barometric pressure and the barometric pressure threshold to determine whether the potential fall event is the fall event; and
output a signal indicating the potential fall event is the fall event based on the change in barometric pressure satisfying the barometric pressure threshold.
8. The device of claim 7 , wherein the barometric pressure threshold is derived from fall event data, and
wherein the confidence estimate values define coefficients of the barometric pressure threshold.
9. The device of claim 7 , wherein the barometric pressure threshold is based on at least one of:
a linear function,
a piecewise linear function, or
a polynomial function.
10. The device of claim 8 , wherein the barometric pressure threshold is derived from a barometric pressure data value corresponding to an event of a person rising from a ground to a standing position.
11. The device of claim 7 , wherein a first confidence estimate value corresponds to a first amount of noise included in the signal generated by the barometric pressure sensor, and
wherein a second confidence estimate value corresponds to a second amount of noise included in the signal generated by the barometric pressure sensor.
12. The device of claim 7 , wherein the change in barometric pressure corresponds to a change in altitude of the barometric pressure sensor.
13. A non-transitory computer-readable medium storing instructions, the instructions comprising:
one or more instructions that, when executed by one or more processors of a device, cause the one or more processors to:
store, for detecting a fall event, barometric pressure data samples in a first buffer,
the device comprising an accelerometer, and
the barometric pressure data samples being barometric pressure data sampled from a signal generated by a barometric pressure sensor;
store, in a second buffer, acceleration data samples,
the acceleration data samples being acceleration data sampled from a signal generated by the accelerometer,
each acceleration data sample corresponding to a respective barometric pressure data sample, and
each acceleration data sample being substantially temporally sampled with the respective barometric pressure data sample;
determine that a potential fall event has occurred based on a magnitude of a particular acceleration data sample exceeding a predetermined acceleration magnitude criterion;
select first barometric pressure data samples in the first buffer having a time of sampling within a predetermine period of time of a sampling of the particular acceleration data sample;
evaluate the first barometric pressure data samples to determine confidence estimate values associated with the first barometric pressure data samples;
output the confidence estimate values;
calculate a barometric pressure threshold based on the confidence estimate values;
select second barometric pressure data samples having a time of sampling during a period of time after the sampling of the particular acceleration data sample;
determine, based on selecting the second barometric pressure data samples, a change in barometric pressure that occurred between a time before the potential fall event and a time after the potential fall event;
compare the change in barometric pressure and the barometric pressure threshold to determine whether the potential fall event is the fall event; and
output a signal indicating the potential fall event is the fall event based on the change in barometric pressure satisfying the barometric pressure threshold.
14. The non-transitory computer-readable medium of claim 13 , wherein the acceleration data samples are further generated by at least one of:
a gyroscope,
a manometer, or
a proximity sensor.
15. The non-transitory computer-readable medium of claim 13 , wherein a confidence estimate value, of the confidence estimate values, is a standard deviation of the first barometric pressure data samples in the first buffer.
16. The non-transitory computer-readable medium of claim 13 , wherein the one or more instructions, that cause the one or more processors to determine that the potential fall event has occurred based on the magnitude of the particular acceleration data sample exceeding the predetermined acceleration magnitude criterion, cause the one or more processors to at least one of:
determine changes in orientation of the device,
determine a free-fall based on acceleration magnitude signals,
determine an indication in accelerometer data that indicate an impact with a ground,
determine a duration of the indication,
determine a duration of a period of free-fall,
determine a standard deviation of a portion of information associated with one or more of an X-axis, a Y-axis, a Z-axis, or magnitude buffers,
determine a mean of the portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determine an area of a spike portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determine an area of a free-fall portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determine a rate of change of the spike portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determine a rate of change of the free-fall portion of the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determine a mode of data in the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determine a median of data in the information associated with one or more of the X-axis, the Y-axis, the Z-axis, or the magnitude buffers,
determine a minimum of data in one or more X-axis, Y-axis, Z-axis, or magnitude threshold values being met, or
determine a peak X-axis, Y-axis, Z-axis, or magnitude threshold value being met.
17. The non-transitory computer-readable medium of claim 13 , wherein the barometric pressure threshold is derived from fall event data, and
wherein the confidence estimate values define coefficients of the barometric pressure threshold.
18. The non-transitory computer-readable medium of claim 13 , wherein the barometric pressure threshold is based on at least one of:
a linear function,
a piecewise linear function, or
a polynomial function.
19. The non-transitory computer-readable medium of claim 13 wherein the barometric pressure threshold is derived from a barometric pressure data value corresponding to an event of a person rising from a ground to a standing position.
20. The non-transitory computer-readable medium of claim 13 , wherein the confidence estimate values are associated with an amount of noise included in the signal generated by the barometric pressure sensor.