System and method for detecting human falls

Abstract

The present invention relates to a system for detecting a person's fall. The system includes a stationary sensor located at a fixed position relative to the person and a wearable sensor worn by the person, and both sensors can report the person's fall independently of each other. Further, the system includes a line-of-sight detection unit configured to determine whether the wearable sensor is within the line of sight of the stationary sensor. To detect a person's fall, the fall detection unit of the system considers a fall report provided by one or both of the stationary sensor and the wearable sensor. Further, to detect a fall, the fall detection unit considers the absence or presence of a line of sight between the wearable sensor and the stationary sensor. Thereby, it is possible to enhance the reliability of detecting a fall, especially when only one of the wearable sensor and the stationary sensor reports a fall.

Description

Technical Field

[0001] The present invention relates to a system and method for detecting a person's fall. Further, the present invention relates to a computer program for detecting a person's fall, and a non-transitory computer-readable data medium storing the computer program.

Background Art

[0002] Since falls are one of the major health threats to the elderly, fall detection systems that can detect a person's fall and provide an alarm accordingly are often used. For example, "Wearable Fall Detection System Using Barometric Pressure Sensors and Machine Learning" by Y. Sun et al., SENSORCOMM 2019: The Thirteenth International Conference on Sensor Technologies and Applications, ISBN: 978-1-61208-744-3 addresses the problem of detecting short falls such as an elderly person falling from a bed or chair when trying to get up. To reduce the number of false alarms, Y. Sun et al. propose using a first pressure sensor attached to a belt and a second pressure sensor attached to a person's shoe, with the second pressure sensor being used as a reference sensor.

[0003] Nevertheless, it is desired to further improve the reliability of systems and methods for detecting a person's fall.

[0004] LI HAOBO et al. disclose preliminary results on multi-sensory recognition of indoor daily activities and fall detection for monitoring the well-being of the elderly at risk of physical and cognitive chronic health disorders. Five different sensors, continuous wave (CW) radar, frequency-modulated CW (FMCW) radar, and an inertial measurement unit including an accelerometer, a gyroscope, and a magnetometer were used to simultaneously collect data from 20 subjects performing 10 activities. It is efficient and economical to select some of them to maximize the classification accuracy and avoid unnecessary calculations for processing insignificant information rather than using all available sensors. Each individual sensor and combinations of several sensors are learned with a quadratic kernel support vector machine classifier.

Summary of the Invention

Problems to be Solved by the Invention

[0005] The present invention is based on the object of providing a system, a method, and a computer program for detecting a person's fall. In particular, with the present system, the present method, and the present computer program, it is possible to detect a person's fall with reliability and at the same time reduce the number of false alarms. As a result, with the present system, the present method, and the present computer program, the amount of false positive alarms or false negative alarms is relatively small, and genuine falls are detected with high accuracy.

Means for Solving the Problems

[0006] According to the present invention, there is provided a system for detecting a person's fall. The system includes a stationary sensor, a wearable sensor, a line-of-sight detection unit, a fall detection unit, and an output unit.

[0007] The stationary sensor is configured to provide a first sensor signal indicating a person's fall. The wearable sensor is configured to provide a second sensor signal indicating a person's fall. The line-of-sight detection unit is configured to determine whether the wearable sensor is within the line-of-sight of the stationary sensor and to provide a line-of-sight signal indicating whether the wearable sensor is within the line-of-sight of the stationary sensor. The fall detection unit is configured to detect a person's fall based on a) the line-of-sight signal and b) the first sensor signal and / or the second sensor signal. Further, the output unit is configured to provide an output signal indicating whether a person's fall has been detected.

[0008] The present invention is based on the recognition that it may be advantageous to use a combination of a wearable sensor and a stationary sensor to detect a person's fall. Since the wearable sensor is always with the person, it has the advantage that it can detect a fall wherever the person goes, i.e., even outdoors. On the other hand, especially when worn on the wrist, the wearable sensor may provide a relatively large number of false positive alarms because certain movements other than a true fall may be detected as a fall. The stationary sensor can detect a person's fall within the field of view of the stationary sensor. When combining the wearable sensor and the stationary sensor in a hybrid fall detection system, it is possible to benefit from the individual advantages associated with each of the stationary sensor and the wearable sensor. Thereby, it is possible to enhance the reliability of fall detection.

[0009] Nevertheless, in a hybrid fall detection system, there may be a situation where only one of the wearable sensor and the stationary sensor detects a fall. In such a situation, it must be determined whether a true fall has occurred or a false positive alarm has been triggered.

[0010] In the system according to the present invention, a line-of-sight detection unit is employed that determines whether the wearable sensor is within the line of sight of the stationary sensor. This can increase the reliability of detecting a true fall while reducing the amount of false positive alarms and / or false negative alarms. The presence or absence of the line of sight may be used as an additional parameter for evaluating whether a true fall has actually occurred. Considering that the wearable sensor and the stationary sensor report a fall and that the wearable sensor is within the line of sight of the stationary sensor, it is possible to more reliably determine that a true fall has actually occurred. Nevertheless, if only one of the wearable sensor and the stationary sensor reports a fall and the wearable sensor is within the line of sight of the stationary sensor, the sensor that did not report the fall should have been in a situation where the fall should have been actually detected, so it may be determined as a false positive alarm. However, if only one of the wearable sensor and the stationary sensor reports a fall and the wearable sensor is not within the line of sight of the stationary sensor, since the other sensor that did not report the fall may not be able to actually detect the fall, it may be determined as a true fall.

[0011] Thus, the presence or absence of the line of sight can be used as an additional parameter for evaluating and thereby increasing the reliability of fall detection performed by the system. Thus, with this system, it is possible to detect a true fall with relatively high accuracy while reducing the amount of false positive alarms and / or false negative alarms.

[0012] In the system, the output signal provided by the output unit may indicate a certain probability that a fall has occurred. For example, referring to the situation described above, when the wearable sensor and the stationary sensor report a fall and the wearable sensor is within the line of sight of the stationary sensor, the output signal may indicate a higher probability that a fall has occurred than in the situation where only one of the wearable sensor and the stationary sensor reports a fall and the wearable sensor is within the line of sight of the stationary sensor. This makes it possible to more accurately adapt further strategies for elderly care to the actual risk of a person falling.

[0013] The wearable sensor may preferably be worn by a person. The wearable sensor may be an accelerometer or may include an accelerometer, and the second sensor signal may represent an acceleration indicating a person's fall. Additionally or alternatively, the wearable sensor may be a speed sensor or may include a speed sensor, and the second sensor signal may represent a speed indicating a person's fall. Additionally or alternatively, the wearable sensor may be a position sensor or may include a position sensor, and the second sensor signal may represent a change in a person's position indicating a person's fall. Additionally or alternatively, the wearable sensor may be a barometric pressure sensor or may include a barometric pressure sensor, and the second sensor signal may represent a change in height indicating a person's fall. The wearable sensor may provide the second sensor signal when a certain threshold value related to each parameter is exceeded. For example, when the wearable sensor is an accelerometer, the second sensor signal may be provided when an acceleration exceeding a predetermined acceleration threshold value is detected. Alternatively, the wearable sensor may be configured to continuously provide the second sensor signal, for example, after a predetermined time interval. The received second sensor signal may be analyzed to detect a fall, for example, by a fall detection unit.

[0014] The second sensor signal may or may include a timestamp associated with an instance of time of a certain parameter, such as when a predetermined threshold is exceeded. For example, a certain speed is detected at a certain time, and since this exceeds a predetermined speed threshold, it is considered to indicate a fall.

[0015] However, it is also possible for the second sensor signal to represent the temporal evolution or trajectory of a certain parameter. For example, the second sensor signal may represent the detected change in position over time, which may indicate a person's fall. For example, the position may be continuously monitored, and in the fall detection unit, a person's fall may be detected from the development of the person's position.

[0016] The fall detection unit may be a processor or processing unit configured to process the first sensor signal, the second sensor signal, and the line-of-sight signal, or may include such a processor or processing unit. In particular, the processor or processing unit may be or include a digital signal processor, a central processing unit of a computer, a microprocessor, a multi-core processor, and / or a field programmable gate array (FPGA).

[0017] The line-of-site detection unit may be implemented by the same processor or processing unit as the fall detection unit, or may be implemented by an external processor or processing unit. The fall detection unit and the line-of-site detection unit are preferably configured to exchange line-of-site signals, such that the line-of-site detection unit has a transmitter and the fall detection unit has a receiver. Also, the line-of-site detection unit and / or the fall detection unit may include a receiver for receiving the first sensor signal and / or the second sensor signal. Accordingly, the wearable sensor and the stationary sensor preferably include a transmitter for providing the first sensor signal and the second sensor signal. The first sensor signal, the second sensor signal, and the line-of-site signal may be provided wirelessly or wired.

[0018] Preferably, the fall detection unit of the system is configured to detect a person's fall when the line-of-site signal indicates that the wearable sensor is within the line-of-site of the stationary sensor and at least one of the first sensor signal and the second sensor signal indicates a person's fall. Depending on the situation, it may be advantageous for a fall to always be detected when at least one of the wearable sensor and the stationary sensor reports a person's fall. In this case, an alarm may be triggered each time either the wearable sensor or the stationary sensor indicates a person's fall.

[0019] Alternatively, the fall detection unit of the system may be configured to detect a person's fall only when the line-of-site signal indicates that the wearable sensor is within the line-of-site of the stationary sensor and both the first sensor signal and the second sensor signal indicate a person's fall. In this case, the fall can be detected with higher reliability, especially by reducing the amount of false positive alarms.

[0020] Alternatively, the system's fall detection unit may be configured to detect a person's fall if the line-of-site signal indicates that the wearable sensor is not within the line of sight of the stationary sensor and at least one of the first sensor signal and the second sensor signal indicates a person's fall. Such a case may be advantageous in situations where the wearable sensor loses line of sight with the stationary sensor. In this case, it may also be advantageous to rely only on the wearable sensor and the stationary sensor individually such that only one of them needs to report a person's fall in order to detect the fall.

[0021] Preferably, the line-of-sight detection unit is configured to repeatedly determine at predetermined time intervals whether the wearable sensor is within the line of sight of the stationary sensor. For example, the line-of-site signal may be provided after each time interval, or alternatively, may be provided only when a predetermined situation is detected. Such a predetermined situation may be detected using an accelerometer that detects a sudden acceleration, a barometric pressure sensor that detects a sudden change in height, or a position sensor that detects a sudden change in position. The time interval at which the line-of-site signal is provided may be dynamically adapted. For example, if a predetermined situation is detected, the frequency at which the line-of-site signal is provided may be increased. This allows for determining several times during a certain time span, e.g., during a fall, whether the wearable sensor is within the line of sight of the stationary sensor.

[0022] The fall detection unit may be configured to use a single line-of-sight signal associated with one of a predetermined time interval, or may be configured to use a plurality of line-of-sight signals associated with two or more of the predetermined time intervals to detect a fall. Using two or more line-of-sight signals to determine whether a fall has occurred can further enhance the reliability of fall detection. For example, there is a possibility that the line-of-sight is lost at a certain time during a fall. In this case, the fall detection unit may use this situation of losing the line-of-sight to evaluate the reliability of the first and second sensor signals. Therefore, to detect a person's fall, the fall detection unit may be configured to consider the moment of loss of the line-of-sight. If the line-of-sight was present in most of the detected falls, the loss of the line-of-sight may indicate that there is something covering the wearable sensor, such as the person himself or furniture, after the fall.

[0023] Additionally or alternatively, the line-of-sight detection unit may be configured to determine whether the wearable sensor is within the line-of-sight of the stationary sensor when at least one of the first sensor signal and the second sensor signal indicates a person's fall. In this case, the determination of the presence or absence of the line-of-sight between the wearable sensor and the stationary sensor is triggered by an indication of a person's fall by at least one of the first sensor signal and the second sensor signal. Therefore, the line-of-sight detection unit may be configured to determine whether the wearable sensor is within the line-of-sight of the stationary sensor when receiving the first sensor signal and / or the second sensor signal indicating a person's fall.

[0024] To determine whether the wearable sensor is within the line of sight of the stationary sensor, it may be advantageous for the line-of-sight detection unit to be configured to consider at least one stationary object within the field of view of the stationary sensor. A stationary object may be within the wearable sensor and the line of sight of the stationary sensor, thereby potentially preventing the line-of-sight detection unit from detecting the presence of the wearable sensor and the line of sight of the stationary sensor. For example, there may be furniture within the field of view of the stationary sensor, and the line of sight of the wearable sensor and the stationary sensor may be blocked by the stationary object. The line-of-sight detection unit may be configured to provide respective line-of-sight signals indicating that the line of sight is blocked by the stationary object. When receiving a corresponding line-of-sight signal indicating that the line of sight is blocked by the stationary object, the fall detection unit may be configured not to rely on the presence of the line of sight to detect a person's fall. In such a situation, it may be sufficient to detect a person's fall if at least one of the wearable sensor and the stationary sensor reports the person's fall.

[0025] Preferably, the system further includes an external light source operably connected to the fall detection unit and configured to provide an optical signal. The wearable sensor may include an optical sensor configured to detect the optical signal provided by the external light source. The line-of-sight detection unit may be configured to determine that the wearable sensor is within the line of sight of the stationary sensor when the wearable sensor detects the optical signal emitted by the external light source with its optical sensor. For example, the line-of-sight detection unit may be configured to determine whether the wearable sensor is within the line of sight of the stationary sensor when receiving an optical detection signal provided by the optical sensor indicating detection of the optical signal.

[0026] Preferably, the external light source is located near or adjacent to the stationary sensor. When one of the sensors indicates a fall, the external light source may be controlled to emit an optical signal to check whether the wearable sensor is within the line of sight of the stationary sensor in order to evaluate the reliability of the fall detection. For example, the fall detection unit and / or the line of sight detection unit may be configured to control the external light source to emit an optical signal when one of the first sensor signal and the second sensor signal indicates a person's fall. Alternatively or additionally, the fall detection unit and / or the line of sight detection unit may be configured to control the external light source to emit light at a predetermined time interval repeatedly to repeatedly check whether the wearable sensor is within the line of sight of the stationary sensor.

[0027] In addition to or alternatively to the external light source, the stationary sensor may include an integrated light source configured to emit an optical signal. The wearable sensor may include an optical sensor configured to detect the optical signal. The line of sight detection unit may be configured to determine that the wearable sensor is within the line of sight of the stationary sensor when the optical sensor of the wearable sensor detects the optical signal emitted by the integrated light source.

[0028] When using a light source integrated with the stationary sensor in addition to or as an alternative to the external light source, since the optical signal is emitted directly from the position of the stationary sensor, the detection of the presence of the line of sight can be improved. Thus, the optical signal may be emitted directly along the virtual line of sight axis that virtually connects the wearable sensor and the stationary sensor.

[0029] The fall detection unit and / or the line-of-sight detection unit may be configured to control the integrated light source to emit an optical signal when one of the first sensor signal and the second sensor signal indicates a human fall. Alternatively or additionally, the fall detection unit and / or the line-of-sight detection unit may be configured to control the integrated light source to emit light after a predetermined time interval repeatedly to repeatedly check whether the wearable sensor is within the line-of-sight of the stationary sensor.

[0030] The optical signal provided by the external light source and / or the integrated light source may be a flash optical signal, an infrared optical signal, or a coded optical signal. The flash optical signal, the infrared optical signal, or the coded optical signal is preferred for reliably distinguishing the optical signal from the ambient light.

[0031] The system further includes an optional navigation system, and the navigation system includes a plurality of beacons configured to communicate with the wearable sensor to determine the relative position of the wearable sensor in the navigation system. The line-of-sight detection unit may be configured to determine whether the wearable sensor is within the line-of-sight of the stationary sensor based on the tracked position of the wearable sensor and the known position of the stationary sensor.

[0032] In such a system, the beacon can establish a tracking environment in which the wearable sensor moves and exchanges signals with the beacon for position detection. The navigation system may be an indoor navigation system installed in the same room or area as the stationary sensor. The line-of-sight detection unit may be configured to consider stationary objects located within the field of view of the stationary sensor to determine whether the wearable sensor is within the line-of-sight of the stationary sensor based on the position of the wearable sensor tracked by the beacon and the known position of the stationary sensor.

[0033] When using a navigation system that includes a plurality of beacons, the wearable sensor is preferably configured to communicate with the beacons of the navigation system by ultra-wideband (UWB) communication, radio-frequency identification (RFID) communication, near field communication (NFC), Bluetooth® communication, and / or visible light communication (VLC) in order to determine the relative position of the wearable sensor. The wearable sensor may include a UWB tag, an RFID tag, an NFC tag, a Bluetooth tag, and / or a VLC sensor. Further, the wearable sensor may be configured to actively determine its position based on communication with the beacon and provide a position signal to a position detection system for further processing.

[0034] Additionally or alternatively to the navigation system that includes a plurality of beacons, the system may include a position detection system that is located on the wearable sensor and includes at least one position sensor configured to track the position of the wearable sensor relative to a stationary sensor. A line-of-sight detection unit may be configured to determine whether the wearable sensor is within the line of sight of the stationary sensor based on the tracked position of the wearable sensor and the known position of the stationary sensor.

[0035] When a position detection system including at least one position sensor located on a wearable sensor is used, beacons distributed in the tracking area may not be adopted, but the wearable sensor communicates directly with the position detection system. The position sensor may be an active or passive sensor that communicates directly with the position detection system. For example, the position sensor may include an optical sensor such as an LED that emits light to indicate the position of the wearable sensor. The emitted light may be captured by the position detection system. Additionally or alternatively, the position sensor may include one or more sensor coils, and the position detection system may be configured to generate an oscillating magnetic field to induce a voltage signal in the sensor coils. The voltage signal may be analyzed by the position detection system to determine the relative position of the wearable sensor.

[0036] A position detection system including at least one position sensor located on a wearable sensor may be an indoor position detection system installed in the same room or area as a stationary sensor. It may be beneficial to equip the stationary sensor with additional position sensors to determine the position of the stationary sensor in the position detection system. Alternatively, the position of the stationary sensor may be provided to the position detection system in advance. For example, the position of the stationary sensor may be known in advance in the coordinate system of the position detection system to enable relative tracking between the wearable sensor and the stationary sensor. Based on the positions of the stationary sensor and the wearable sensor provided by the position detection system, a line-of-sight detection unit may determine whether the wearable sensor is within the line of sight of the stationary sensor. Thereby, the line-of-sight detection unit may be configured to consider stationary objects located within the field of view of the stationary sensor to determine whether the wearable sensor is within the line of sight of the stationary sensor.

[0037] Preferably, the stationary sensor is or includes a radar. With the radar, the stationary sensor may sense a field of view where a person may be present. The radar may be a monostatic frequency modulated continuous wave (FMCW) radar or a WIFI (registered trademark)-based bi-static radar. The wearable sensor may be or may be included in a smartwatch, a neck-worn sensor, a ring-sensor, a head-worn sensor, an ear-worn sensor, or a wrist-worn sensor.

[0038] The system may optionally be further improved in that each of the stationary sensor and the wearable sensor includes a barometric pressure sensor for detecting the height from the floor based on the absolute barometric pressure. The fall detection unit may be configured to detect a person's fall based on the detected heights of the stationary sensor and the wearable sensor. Thus, the detected heights of the stationary sensor and the wearable sensor may be used as additional parameters for the line of sight to enhance the reliability of detecting a person's fall. For example, when a sudden change in height is detected using the barometric pressure sensor, the fall detection unit may determine a fall even if the line of sight signal indicates a loss of line of sight between the wearable sensor and the stationary sensor. The barometric pressure sensor of the stationary sensor may thereby act as a reference sensor. Further, an Inertial Measurement Unit (IMU) may be included in the wearable sensor.

[0039] The present invention also relates to a method for detecting a person's fall. The method includes receiving a first sensor signal indicating a person's fall from a stationary sensor and / or Receiving a second sensor signal indicative of a person's fall from a wearable sensor; Determining whether the wearable sensor is within the line of sight of the stationary sensor; Providing a line-of-sight signal indicative of whether the wearable sensor is within the line of sight of the stationary sensor; Detecting a person's fall based on a) the line-of-sight signal and b) the first sensor signal and / or the second sensor signal; Providing an output signal indicative of whether a person's fall has been detected; including.

[0040] The method may be implemented using the above-described system for detecting a person's fall.

[0041] The method includes receiving a first sensor signal indicative of a person's fall from a stationary sensor, but may not include receiving a second sensor signal indicative of a person's fall from a wearable sensor. Alternatively, the method may include both receiving a first sensor signal indicative of a person's fall from a stationary sensor and receiving a second sensor signal indicative of a person's fall from a wearable sensor.

[0042] The present invention also relates to a computer program for detecting a person's fall, which, when executed on a computer, includes instructions for performing the steps of the above-described method. The computer program may be stored on a non-transitory computer-readable data medium.

[0043] According to one aspect, an object is to provide a system, a method, and a computer program for monitoring a person. The monitoring may include fall detection. In one example, the system includes a stationary sensor configured to provide a first sensor signal indicating a monitoring state of the person, a wearable sensor configured to provide a second sensor signal indicating a monitoring state of the person, a line-of-sight detection unit configured to determine whether the wearable sensor is within the line of sight of the stationary sensor and to provide a line-of-sight signal indicating whether the wearable sensor is within the line of sight of the stationary sensor, a monitoring detection unit configured to detect the monitoring state of the person based on a) the line-of-sight signal and b) the first sensor signal and / or the second sensor signal, and an output unit (optionally) configured to provide an output signal indicating the monitoring state of the person.

[0044] In this aspect, the monitoring may include detecting a person's fall, vital signs such as a person's heart rate, breathing, or any other health / care-related monitoring of the person, or characteristics such as the person's presence, movement, activity, etc. The monitoring status includes whether a fall has been detected and the output of monitoring such as heart rate, activity, etc. It should be understood that this aspect of the present invention can also be any combination of the dependent claims or the above embodiments with their respective independent claims. For example, the fall detection unit may be interchangeable with the monitoring detection unit.

[0045] It should be understood that the heating system of claim 1, the method of claim 13, the use of the heating element of claim 13, and the computer program of claim 14 have similar and / or the same preferred embodiments, especially the embodiments as described in the dependent claims.

[0046] It should also be understood that the preferred embodiments of the present invention can be any combination of the dependent claims or the above embodiments with their respective independent claims.

[0047] These and other aspects of the present invention will become apparent and will be elucidated with reference to the embodiments described below.

Brief Description of the Drawings

[0048]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Mode for Carrying Out the Invention

[0049] Figure 1 shows a system 100 for detecting a fall of a person 102 (not part of the system). The person 102 wears a smartwatch including a wearable sensor 104. The wearable sensor 104 is on the person 102, such that the fall of the person 102 can be detected wherever the person 102 goes. To detect the fall of the person 102, the wearable sensor 104 includes an accelerometer configured to detect acceleration. The wearable sensor 104 is configured to provide a second sensor signal 106 indicating whether a fall has occurred, for example representing the detected acceleration. In particular, the wearable sensor 104 may be configured to provide the second sensor signal 106 when the detected acceleration exceeds a predetermined acceleration threshold. When the detected acceleration exceeds the predetermined threshold, the wearable sensor 104 thus reports the fall of the person 102 by providing the second sensor signal 106 indicating the fall.

[0050] Alternatively, the second sensor signal 106 may be provided continuously, for example, at predetermined time intervals, and the analysis of the second sensor signal 106 may be performed outside the wearable sensor 104. For example, a fall detection unit 108 may be used to analyze the second sensor signal 106. The analysis of the second sensor signal 106 may include a comparison with a threshold for identifying a fall of the person 102, for example, a predetermined acceleration threshold.

[0051] In addition to or alternatively to the accelerometer, the wearable sensor may include a velocity sensor configured to measure the velocity of the person 102, a position sensor configured to measure a change in the position of the person 102, or a barometric pressure sensor configured to detect a change in the height of the person 102.

[0052] System 100 is a hybrid fall detection system that includes a stationary sensor 110 in addition to the wearable sensor 104. The stationary sensor 110 has the advantage of not providing a positive alarm that wrongly indicates a fall of the person 102 triggered by certain movements of the person 102 that could potentially be considered a fall by the wearable sensor 104. For example, if the wearable sensor is included in a smartwatch, a sudden movement of the arm could be misrecognized as a fall by the wearable sensor. Still, the stationary sensor 110 may have a limited field of view, and the person 102 may walk outside of that field of view such that the stationary sensor 110 cannot detect the person 102's fall.

[0053] The stationary sensor 110 includes a radar that is a monostatic FMCW radar. Alternatively, the stationary sensor 110 may include a Wi-Fi (registered trademark)-based bistatic radar. When the stationary sensor 110 detects a fall of the person 102, the stationary sensor 110 provides a first sensor signal 112 indicating the fall. Alternatively, the stationary sensor may continuously provide the first sensor signal 112, and the analysis of the first sensor signal 112 may be performed by the fall detection unit 108. From the first sensor signal 112, the fall detection unit 108 may extract features indicating a fall of the person 102 and determine whether a true fall of the person 102 has occurred based on the extracted features.

[0054] When the wearable sensor 104 and the stationary sensor 110 are used in combination, there may be a situation where only one of the sensors 104, 110 reports a fall of the person 102. In this case, the system 100 needs to determine whether a true fall has occurred or the fall report is a false positive. To make this determination, it is advantageous to know the spatial relationship between the wearable sensor 104 and the stationary sensor 110. In particular, it may be advantageous to know the location of the wearable sensor 104 relative to the stationary sensor 110. If the wearable sensor 104 is within the line of sight 114 of the stationary sensor 110 and only one of the two sensors 104, 110 reports a fall, the fall report is more likely to be a false positive. Further, if the wearable sensor 104 is not within the line of sight 114 of the stationary sensor 110 and only one of the two sensors 104, 110 reports a fall, it is more likely that an actual true fall has occurred.

[0055] To analyze the spatial relationship between the wearable sensor 104 and the stationary sensor 110, the system 100 includes a line-of-sight detection unit 116 configured to determine whether the wearable sensor 104 is within the line of sight 114 of the stationary sensor 110. Detection of the line of sight 114 can be performed in various ways. For example, as described in detail with reference to FIG. 2, the line of sight 114 is external to or incorporated in the stationary sensor 110 and can be detected using a light source configured to provide a light signal that can be detected by a light sensor included in the wearable sensor 104 when the wearable sensor 104 is within the line of sight 114 of the stationary sensor 110. Additionally or alternatively, as described in detail with respect to FIGS. 3 and 4, a navigation system or a position detection system may be used to track the position of the wearable sensor 104 relative to the stationary sensor 110, respectively.

[0056] The line-of-site detection unit 116 is configured to provide a line-of-site signal 118 indicating whether the wearable sensor 104 is within the line of site 114 of the stationary sensor 110. The line-of-site detection unit 116 may continuously provide the line-of-site signal 118, for example, at a predetermined time interval, and may indicate the absence or presence of the line of site 114 for each time interval. Alternatively, the line-of-site detection unit 116 may provide the line-of-site signal 118 only when at least one of the wearable sensor 104 and the stationary sensor 110 reports a fall of the person 102. For example, the fall detection unit 108 may trigger the line-of-site detection unit 116 to check the presence of the line of site 114 when at least one of the wearable sensor 104 and the stationary sensor 110 reports a fall of the person 102.

[0057] The fall detection unit 108 of the system receives the line-of-site signal 118 from the line-of-site detection unit 116 and may determine whether the person 102 has fallen or not based on the line-of-site signal 118 and a fall of the person 102 reported by at least one of the wearable sensor 104 and the stationary sensor 110. The result of the determination made by the fall detection unit 108 may be provided to an output unit 120 of the system configured to provide an output signal 122 indicating whether a fall of the person 102 has been detected. Based on whether the output signal 122 indicates a fall of the person 102, further actions may be triggered. For example, an alarm may be activated to notify the fall of the person 102. Also, the output signal 122 may represent the probability that a true fall has occurred. Based on the provided probability, further actions may be triggered. For example, if the probability that a true fall has occurred exceeds a predetermined probability threshold, an alarm may be activated accordingly. Also, the output signal 122 may be received by a device having a display for visualizing a message indicating the probability that a true fall has occurred. The user of the device may determine whether further actions are required based on the provided message.

[0058] Figure 2 shows a system 200 for detecting a fall of a person 202 (who is not part of the system). Similar to the system 100 described with reference to FIG. 1, the system 200 includes a wearable sensor 204. Similar to the wearable sensor 104, the wearable sensor 204 may be included in a smartwatch or a similar device worn by the person 202. The wearable sensor 204 may be implanted subcutaneously in the person 202. The wearable sensor 204 includes a barometric pressure sensor 206 for detecting the height from the floor based on the absolute atmospheric pressure. A change in the height from the floor detected by the barometric pressure sensor 206 may indicate a fall of the person 202. The wearable sensor 204 is configured to provide, for example, a second sensor signal 208 indicating a fall of the person 204, representing the height from the floor detected by the barometric pressure sensor 206. In addition to the barometric pressure sensor 206, the wearable sensor 204 may include an accelerometer, a speed sensor, or a position sensor for detecting a fall of the person 202.

[0059] In addition to the barometric pressure sensor 206, the wearable sensor 204 includes an optical sensor 210 configured to detect an optical signal 212 to determine whether the wearable sensor 204 is within the line of sight 214 of the stationary sensor 216 of the system. The optical sensor 210 may be a semiconductor-based photodetector such as a photodiode or a phototransistor. When the optical sensor 210 detects the optical signal 212, the wearable sensor 204 may provide an optical detection signal 218 indicating whether the optical signal 212 has been detected. The optical signal 212 may be visible flash light or an infrared signal or coded light.

[0060] In the operation of system 200, the optical signal 212 may preferably be provided by an external light source 220 located at or at least near the stationary sensor 216. Instead of the external light source 220, an integrated light source incorporated into the stationary sensor 216 may be used. It is particularly preferred that the external light source 220 or alternatively the integrated light source is located at the same position as the stationary sensor 216 or at least near the stationary sensor 216. Thereby, when the optical sensor 210 detects the optical signal 212 from the external or integrated light source, the wearable sensor 204 can be assumed to be within the line of sight 214 of the stationary sensor 216. Such an assumption can be valid since the optical signal 212 travels substantially along the line of sight axis that virtually connects the wearable sensor 204 and the stationary sensor 216.

[0061] Thus, the detection of the optical signal 212 by the optical sensor 210 may be used by the line of sight detection unit 222 of the system to positively determine the absence or presence of the line of sight 214. If the line of sight detection unit 222 receives an optical detection signal 218 from the wearable sensor 200 indicating the presence of the line of sight 214, the line of sight detection unit 222 may provide a line of sight signal 224 indicating the presence of the line of sight 214 between the wearable sensor 204 and the stationary sensor 216.

[0062] The stationary sensor 216 includes a radar which is a Wi-Fi (registered trademark)-based bistatic radar. Alternatively, the radar may be a monostatic FMCW radar or the like. When the stationary sensor detects the fall of the person 202, the stationary sensor 216 may provide a first sensor signal 228 indicating the fall. Also, the first sensor signal 228 may be provided repeatedly after a predetermined time interval, and the analysis of the first sensor signal 228 may be performed externally, for example, by a fall detection unit 230 of the system. In addition to the radar, the stationary sensor 216 includes a pressure sensor 226 that serves as a reference sensor for the pressure sensor 206 included in the wearable sensor 204. Thus, the height from the floor detected by the pressure sensor 206 of the wearable sensor 204 may be compared with the height detected by the pressure sensor 226 of the stationary sensor 216. For example, a change in the height of the wearable sensor 204 with respect to the reference height provided by the pressure sensor 226 of the stationary sensor 216 may be determined.

[0063] In addition to the line-of-sight signal 224, the fall detection unit 230 is configured to detect a fall of the person 202 using the first sensor signal 228 and / or the second sensor signal 208. Further, the fall detection unit 230 may use the reference height measured by the barometric pressure sensor 226 of the stationary sensor 216. In particular, when receiving the first sensor signal 226 or the second sensor signal 208, the fall detection unit 230 may trigger the line-of-sight detection unit 222 to provide a line-of-sight signal 224 indicating the absence or presence of the line-of-sight 214 between the wearable sensor 204 and the stationary sensor 216. When triggered to provide the line-of-sight signal 224, the line-of-sight detection unit 222 may trigger the external light source 220 to emit an optical signal 212. If the optical sensor 210 detects the optical signal 212 and the wearable sensor 204 provides respective optical detection signals 218 to the line-of-sight detection unit 222, the line-of-sight detection unit 222 provides the line-of-sight signal 224 indicating the presence of the line-of-sight 214 to the fall detection unit 230. In this case, the fall detection unit 230 may require both the wearable sensor 204 and the stationary sensor 216 to report a fall of the person 200 in order to detect a true fall. If no optical detection signal is received by the line-of-sight detection unit, after a predetermined time span, the line-of-sight detection unit 224 may provide a line-of-sight signal indicating the absence of the line-of-sight 214. For example, if the line-of-sight detection unit 224 reports the absence of the line-of-sight 214, the fall detection unit 230 may already positively determine a fall of the person 200 if only one of the wearable sensor 204 and the stationary sensor 216 reports a fall. Based on the determination of whether a fall has occurred provided by the fall detection unit 230, the output unit 232 of the system may provide an output signal 234 indicating whether a fall has been detected.

[0064] Figure 3 shows a system 300 for detecting the fall of a person 302 (who is not part of the system). The system 300 includes an indoor navigation system 304 that engages in determining the relative positions of a plurality of beacons 306 for determining the relative position of a wearable sensor 308 worn by the person 302. The indoor navigation system 304 is used by a line-of-sight detection unit 312 to determine the absence or presence of a line of sight 310 between the wearable sensor 308 and a stationary sensor 314. The indoor navigation system 304 may be used in conjunction with the system 100 described with reference to FIG. 1. Also, the navigation system 304 may be used additionally or alternatively for an external or integrated light source for emitting an optical signal that can be captured by an optical sensor included in the wearable sensor described with respect to FIG. 2.

[0065] The beacons 306 are distributed across the area where the person 302 moves. For example, the beacons 306 may be attached to walls, ceilings, furniture, etc. The beacons 306 are configured to emit a beacon signal 316 that can be received by the wearable sensor 308. For example, the wearable sensor 308 may include a tag configured to detect the beacon signal 316 provided by the beacon 306. The beacons 306 are configured to communicate via Bluetooth (registered trademark) and may be Bluetooth low energy (BLE) beacons. Alternatively, the beacons 306 may be configured to communicate by UWB communication, RFID communication, NFC communication, or VLC communication. Accordingly, the tag of the wearable sensor configured to communicate with the beacon 306 is a Bluetooth tag, but may be a UWB tag, RFID tag, NFC tag, or VLC tag when using different communication protocols.

[0066] The wearable sensor 308 may be configured to determine its position based on communication with the beacon 306 and provide a position signal to the indoor navigation system 304. Alternatively, the wearable sensor 308 may be configured to preferably wirelessly transmit position data based on communication with a beacon, and the indoor navigation system 304 may be configured to determine the position of the wearable sensor based on the provided position data. The indoor navigation system 304 may track the position of the wearable sensor 308 and provide the tracked position 318 to the off-site detection unit 312 of the system. The relative position of the stationary sensor 314 may be provided in advance to the off-site detection unit 312, or may be determined by the indoor navigation system 304 and provided to the off-site detection unit 312. Based on the tracked position 318 and the known position of the stationary sensor 314, the off-site detection unit 312 may determine whether the wearable sensor 308 is within the off-site 310 of the stationary sensor 314. The off-site detection unit 312 may provide an off-site signal 320 indicating the absence or presence of the off-site 310.

[0067] The wearable sensor 308 of the system 300 includes a speed sensor and is configured to report a fall of the person 302 when the speed sensor senses a speed exceeding a predetermined speed threshold. Additionally or alternatively to the speed sensor, the wearable sensor 308 may include a barometric pressure sensor as described with reference to FIG. 2, or an accelerometer included in the wearable sensor of the system 100 as described with respect to FIG. 1. Additionally or alternatively, the wearable sensor 308 may include a position sensor. When the speed sensor detects a speed exceeding a predetermined threshold, or alternatively, when the speed detected at repeated predetermined time intervals is provided, the wearable sensor 308 may provide a second sensor signal 322 indicating a fall of the person 302.

[0068] The stationary sensor 314 may include a monostatic FMCV radar, as described with reference to, for example, FIG. 1, or a radar that may be a Wi-Fi (registered trademark)-based bistatic radar, as described with reference to, for example, FIG. 2. When detecting a fall, or continuously at a predetermined time interval, the stationary sensor 314 may provide a first sensor signal 324 indicating a fall of the person 302.

[0069] Based on at least one of the first sensor signal 324 and the second sensor signal 322, and using the line-of-sight signal 320, the fall detection unit 328 of the system is configured to detect a fall of the person 302, as described, for example, with respect to the fall detection unit of the system 100 or the system 200 described with reference to FIGS. 1 and 2 respectively. The output unit 330 of the system may provide an output signal 332 indicating whether a fall has been detected by the fall detection unit 328.

[0070] FIG. 4 shows a system 400 for detecting a fall of a person 402 (not part of the system). The system 400 includes a position detection system 406 configured to provide position information 408 representing the relative position of the wearable sensor 404 to determine whether the wearable sensor 404 is within the line of sight 410 of the stationary sensor 412. The position detection system 406 may be used additionally or alternatively to an indoor navigation system, as described with reference to, for example, FIG. 3. Also, the position detection system 406 may be used additionally or alternatively to an external or integrated light source configured to emit an optical signal to be detected by the optical sensor of the wearable sensor described with reference to FIG. 2. Also, the position detection system 406 may be used in combination with the system 100 described with reference to FIG. 1.

[0071] The position detection system 406 is an electromagnetic position detection system that includes a magnetic field generator configured to generate an oscillating magnetic field 416. The wearable sensor 404 includes a position sensor 414 that includes at least one sensor coil. The oscillating magnetic field 416 can induce a voltage in the sensor coil of the position sensor 414, and the wearable sensor 404 may provide a position detection signal representative of the induced voltage. Since the induced voltage depends on the distance to the magnetic field generator and also on the orientation of the sensor coil within the oscillating magnetic field, the relative position of the wearable sensor may be determined from the induced voltage. Based on the position detection signal provided by the position sensor, the position detection system 406 may determine the relative position of the wearable sensor 404. Optionally, the stationary sensor 412 may similarly be provided with a position sensor so that the position detection system 406 can determine the relative position of the stationary sensor 412. Alternatively, the relative position of the stationary sensor may be known in advance, for example, as coordinates in the coordinate system of the position detection system. Alternatively to the electromagnetic position detection system, the position detection system 406 may be an optical position detection system, the position sensor 414 may include one or more light emitting diodes (LEDs) that emit light, etc., and based on the light, the position detection system 406 may determine the relative position of the wearable sensor 404. Alternatively, the position sensor 414 may include a photodetector for detecting light provided by one or more light sources used for position detection.

[0072] In particular, when using an optical position detection system, it may be possible to employ a plurality of modulated LEDs as light sources distributed throughout a room or an area where a person moves. For example, the LEDs may be part of one or more lighting fixtures, or may be constituted by them, or may be attached to one or more main power sockets, etc. A person may wear a position sensor, for example, as part of a wearable sensor, for example, on the person's wrist. The position sensor may sense the modulated light emitted by the distributed LEDs with a photodetector in order to determine the relative position of the position sensor from the detected modulation of the light. For example, the position sensor may be constituted by a combination of segmented photodiodes and apertures, and may be combined with processing electronics configured to recognize the modulation of the light emission of the distributed LEDs. Thereby, it is possible to detect the position of the wearable sensor, and thus it can be used to determine whether the wearable sensor is within the line of sight of the stationary sensor.

[0073] Alternatively, it may be possible to employ a plurality of position sensors distributed throughout a room or an area where a person moves. Preferably, the relative positions of the distributed position sensors are known. For example, the position sensors may be attached to lighting fixtures and / or main sockets, etc., or may be part of them. In this case, the modulated LED may be worn on a person, preferably as part of the wearable sensor, so as to emit modulated light that can be detected by the plurality of distributed position sensors in order to detect the position of the wearable sensor relative to the distributed position sensors. It may be possible to determine whether the wearable sensor is within the line of sight of the stationary sensor from the relative position of the modulated LED worn on the person.

[0074] In addition to the position sensor 414, the wearable sensor 404 may include a barometric pressure sensor, such as described with reference to FIG. 2, or an accelerometer, such as described with reference to FIG. 2, or a speed sensor, such as described with reference to FIG. 3, to detect a fall of the person 402. Further, the wearable sensor 404 is configured to provide a second sensor signal 418 indicating a fall of the person 402. The stationary sensor includes a radar that can be a monostatic FMCW radar or a Wi-Fi (registered trademark)-based bistatic radar. When detecting a fall of the person 402, the stationary sensor 412 provides a first sensor signal 420 indicating the fall.

[0075] Based on the position information 408 representing the relative position of the wearable sensor 404 tracked by the position detection system 406 and the known relative position of the stationary sensor 412, the line-of-sight detection unit 422 of the system may determine whether the wearable sensor 404 is within the line of sight 410 of the stationary sensor 412. When checking the line of sight 410, the line-of-sight detection unit 422 may provide a line-of-sight detection signal 424 indicating the presence or absence of the line of sight.

[0076] The fall detection unit 426 of the system may use the provided line-of-sight signal 424 and at least one of the first sensor signal 420 and the second sensor signal 418 indicating a fall of the person 402 to determine the likelihood that a true fall of the person 402 has occurred. The result of the fall detection performed by the fall detection unit 426 may be transferred to an output unit 428 configured to provide an output signal 430 indicating the result.

[0077] FIG. 5 shows a flowchart diagram representing a method for detecting a person's fall. The method described below may be implemented using, for example, the system 100 described with reference to FIG. 1, the system 200 described with reference to FIG. 2, the system 300 described with reference to FIG. 3, or the system 400 described with reference to FIG. 4.

[0078] In particular, the method may be implemented by a computer program comprising instructions for performing the method steps when executed on a computer. The computer may be part of system 100, system 200, system 300, or system 400. For example, the computer may include / implement the functionality of the line-of-site detection unit and / or the fall detection unit and / or the output unit as described above with respect to systems 100, 200, 300, and / or 400.

[0079] In this method, a first sensor signal indicating a person's fall is preferably received from a stationary sensor including a radar (step S1). Additionally or alternatively to receiving the first sensor signal, in this method, a second sensor signal indicating a person's fall is received from a wearable sensor, which may be part of a smartwatch or the like worn by the person (step S2). Additionally to receiving the first sensor signal and / or the second sensor signal, it is determined whether the wearable sensor is within the line of sight of the stationary sensor (step S3). Determining whether the wearable sensor is within the line of sight of the stationary sensor can be accomplished using an external or integrated light source for emitting an optical signal and the optical sensor of the wearable sensor, as described with respect to FIG. 2. Alternatively or additionally, the line of sight may be determined using a navigation system as described with respect to FIG. 3 and / or by employing a position detection system as described with reference to FIG. 4.

[0080] When it is determined whether the wearable sensor is within the line of sight of the stationary sensor, a line-of-sight signal indicating the absence or presence of the line of sight is provided (step S4). Using the line-of-sight signal and the first sensor signal and / or the second sensor signal, a person's fall is detected, for example, using a fall detection unit, as described with reference to FIGS. 1, 2, 3, or 4 (step S5). Thereafter, an output signal indicating whether a person's fall has been detected is provided (step S6).

[0081] In summary, the present invention relates to a system for detecting a person's fall. The system includes a stationary sensor located at a fixed position relative to the person and a wearable sensor worn by the person, and both sensors can report the person's fall independently of each other. Further, the system includes a line-of-sight detection unit configured to determine whether the wearable sensor is within the line of sight of the stationary sensor. To detect a person's fall, the fall detection unit of the system considers the fall report provided by one or both of the stationary sensor and the wearable sensor. Further, to detect a fall, the fall detection unit considers the absence or presence of the line of sight between the wearable sensor and the stationary sensor. Thereby, it is possible to enhance the reliability of detecting a fall, especially when only one of the wearable sensor and the stationary sensor reports a fall.

[0082] Other variations to the disclosed embodiments can be understood by those skilled in the art upon review of the drawings, the present disclosure, and the appended claims, and can be carried out when practicing the claimed invention.

[0083] In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite articles "a" or "an" do not exclude a plurality.

[0084] A single unit or device may perform the functions of several items recited in the claims. The mere fact that certain means are recited in mutually different dependent claims does not indicate that a combination of these means cannot be used advantageously.

[0085] Procedures such as receiving first and second sensor signals, determining a line of sight, detecting a person's fall, etc., which are executed by one or more units or devices, can be executed by any other number of units or devices. Detecting a person's fall in accordance with these procedures, in particular, a method for detecting a person's fall implemented by a fall detection system, can be implemented as program code means of a computer program and / or as dedicated hardware.

[0086] The computer program may be stored / distributed in a suitable medium, such as an optical storage medium or a solid state medium, which is supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

[0087] Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A system for detecting when a person falls, said system A stationary sensor configured to provide a first sensor signal indicating a person has fallen, A wearable sensor configured to provide a second sensor signal indicating a person has fallen, A line-of-sight detection unit configured to determine whether the wearable sensor is within the line of sight of the stationary sensor, and to provide a line-of-sight signal indicating whether the wearable sensor is within the line of sight of the stationary sensor, a) a fall detection unit configured to detect a person falling based on the line of sight signal and b) the first sensor signal and / or the second sensor signal, An output unit configured to provide an output signal indicating whether a person has fallen has been detected, A system that includes this.

2. The fall detection unit is configured to detect a person falling when the line of sight signal indicates that the wearable sensor is within the line of sight of the stationary sensor, and at least one of the first sensor signal and the second sensor signal indicates a person falling.

3. The system according to claim 1 or 2, wherein the line of sight detection unit is configured to repeatedly determine at predetermined time intervals whether the wearable sensor is within the line of sight of the stationary sensor.

4. The system according to claim 1 or 2, wherein the line-of-sight detection unit is configured to determine whether the wearable sensor is within the line of sight of the stationary sensor when at least one of the first sensor signal and the second sensor signal indicates a person falling.

5. The system according to claim 1 or 2, wherein the line of sight detection unit is configured to consider at least one stationary object within the field of view of the stationary sensor in order to determine whether the wearable sensor is within the line of sight of the stationary sensor.

6. The system according to claim 1 or 2, wherein the system includes an external light source operably connected to the fall detection unit and located near the stationary sensor, the wearable sensor includes a light sensor configured to detect an optical signal, and the line of sight detection unit is configured to determine that the wearable sensor is within the line of sight of the stationary sensor when the wearable sensor detects an optical signal emitted by the external light source with its light sensor.

7. The system according to claim 1 or 2, wherein the stationary sensor includes an integrated light source, the wearable sensor includes a light sensor configured to detect an optical signal, and the line of sight detection unit is configured to determine that the wearable sensor is within the line of sight of the stationary sensor when the wearable sensor detects an optical signal emitted by the integrated light source with its light sensor.

8. The system according to claim 1 or 2, wherein the system includes a navigation system, the navigation system includes a plurality of beacons configured to communicate with the wearable sensor in order to determine the relative position of the wearable sensor in the navigation system, and the line of sight detection unit is configured to determine whether the wearable sensor is in the line of sight of the stationary sensor based on the tracked position of the wearable sensor and the known position of the stationary sensor.

9. The system according to claim 8, wherein the wearable sensor is configured to communicate with a beacon of the navigation system by ultra-wideband communication, radio frequency identification communication, near-field communication, Bluetooth® communication, or visible light communication.

10. The system according to claim 1 or 2, wherein the system includes a position detection system, the position detection system includes at least one position sensor located on the wearable sensor and configured to track the position of the wearable sensor relative to the stationary sensor by the position detection system, and the line of sight detection unit is configured to determine whether the wearable sensor is in the line of sight of the stationary sensor based on the tracked position of the wearable sensor and the known position of the stationary sensor.

11. The system according to claim 1 or 2, wherein the stationary sensor is or includes radar.

12. The system according to claim 1 or 2, wherein each of the stationary sensor and the wearable sensor includes a barometric pressure sensor for detecting height from the floor based on absolute pressure, and the fall detection unit is configured to detect a person falling based on the height detected by the stationary sensor and the wearable sensor.

13. A computer-based method for detecting a person falling, the method is: A first sensor signal indicating a person falling is received from a stationary sensor, and / or Receiving a second sensor signal from a wearable sensor indicating a person has fallen, To determine whether the wearable sensor is within the line of sight of the stationary sensor, To provide a line-of-sight signal indicating whether the wearable sensor is within the line of sight of the stationary sensor, a) detecting a person falling based on the line-of-sight signal and b) the first sensor signal and / or the second sensor signal, To provide an output signal indicating whether a person has fallen, Methods that include...

14. A computer program for detecting a person falling, which, when executed on a computer, includes instructions for performing the steps of the method according to claim 13.

15. A non-temporary computer-readable data medium storing the computer program described in claim 14.

Images