A fall and similar action warning method based on inertial sensing

Abstract

The present invention discloses a fall and similar action warning method based on inertial sensing. The method comprises: obtaining a three-axis acceleration signal and a three-axis angular velocity signal of a target in a body coordinate system, and calculating a combined acceleration; obtaining a three-axis acceleration signal and a three-axis angular velocity signal of a world coordinate system according to a mapping relationship between the body coordinate system and the world coordinate system; calculating a three-axis motion acceleration in the world coordinate system; judging whether the target is in a weightless state by comparing the combined acceleration with a set threshold; and in the case where the target is judged to be in a weightless state, performing a fall warning according to a change in an instantaneous displacement in a vertical direction and / or an inclination angle, wherein the instantaneous displacement is obtained based on a motion speed and motion acceleration of the target in a vertical direction, and the change in the inclination angle is obtained based on an angular velocity signal of an X-axis and a Y-axis in a world coordinate system. The present invention improves the accuracy of fall prediction.

Description

Technical Field

[0001] The present invention relates to the technical field of human fall detection, and more specifically, to a fall and similar action warning method based on inertial sensing. Background Art

[0002] Studies have shown that falling is the most common aging syndrome and the main cause of injury for people over 65 years old. At present, the solutions for judging falling mainly include fall detection based on video images, fall detection based on audio signals, and fall detection based on wearable devices. In the solution based on video images, the real-time movement of the object is monitored by the camera, and then the action category in the image is identified, the current action is matched with the preset fall action, and it is determined whether the fall has occurred. This solution cannot guarantee the privacy and security of users, can detect fewer fall behaviors, and has limited usage scenarios and high costs. In the solution based on audio signals, the fall event is determined by analyzing the frequency part of the vibration caused by the impact, but the installation of such equipment is relatively complicated and the capital investment is relatively large. For the solution based on wearable devices, the sensor module of the common "fall alarm" generally uses a single three-axis accelerometer, the measurement accuracy and data mode are limited, or the threshold algorithm used is not good enough. The biggest defect is that it can only realize the fall alarm, that is, the alarm and help after the fall occurs, but it cannot predict the fall and remind the user in advance.

[0003] In the prior art, patent application CN202110819148.8 discloses a fall warning system based on neural network signals, including a detachable sole that can be installed on the bottom of flat shoes and a posture adjustment device arranged in the sole; the detachable sole includes a sole body in the shape of a groove, and a cross plate is arranged in the middle of the sole body, and the cross plate separates the sole body into a connection cavity located at the upper part for supporting flat shoes and an installation cavity located at the lower part; a cross support plate is arranged in the installation cavity, and two strip-shaped installation ports are symmetrically arranged on the left and right sides of the front and rear ends of the cross support plate, and an action port is arranged at the bottom of the sole body corresponding to each installation port. This scheme reduces the probability of falling by mechanical means and physical methods, but is ineffective for non-daily walking actions in life (for example, falling out of bed), and cannot make a fall warning.

[0004] Patent application CN201910865163.9 discloses a medical anti-fall warning device, including a tightening ring, a connecting block fixedly connected to one side of the tightening ring, a box body fixedly connected to one side of the connecting block, a buffer block fixedly connected to the inner side of the tightening ring, and a connecting assembly is provided on the tightening ring, a plurality of vents are provided on the tightening ring, and a battery and an alarm are fixedly connected to the top of the inner wall of the box body. This scheme uses the size of the human body's tilt angle to determine whether a fall has occurred, but it cannot identify and judge normal large tilt angle behaviors such as bending over and lying down, and the probability of false alarm is relatively high.

[0005] Patent application CN201810134106.9 discloses a big data system for warning and positioning falls of the elderly, including a wearable signal collection device and a cloud data analysis system; the wearable signal collection device includes a first pressure collection unit, a second pressure collection unit and a third pressure collection unit for collecting pressure signals at the heel, second metatarsal, phalanges and first metatarsal, phalanges of the user when walking; the first pressure collection unit, the second pressure collection unit and the third pressure collection unit are all connected to a local storage module, the local storage module is connected to the user's mobile phone through a short-distance wireless communication unit, and the mobile phone is remotely connected to the cloud data analysis system. Although this solution can effectively protect the personal safety of the elderly and reduce the risk of falling of the elderly to a certain extent, it is poor in the effect of warning falls. For example, when the human body jumps up and down, lies down in place, etc., it is impossible to distinguish and predict falls.

[0006] In summary, the existing fall detection solutions have the following defects:

[0007] 1) Using video signal acquisition equipment to collect human behavior video information, and realizing the reverse prediction solution by processing human behavior video requires the assistance of video acquisition equipment, and can only work in specific areas covered by video acquisition equipment. Moreover, the image processing method can only have a high recognition rate for typical human fall behaviors, and there are blind spots for recognition. At the same time, this type of solution requires more equipment, the hardware device is bulky, the installation and use are cumbersome, the cost is also high, and it is not convenient to use in common life scenarios.

[0008] 2) The biomechanical-based fall warning system for the elderly uses pressure sensors to collect physiological pressure signals from the soles of the feet to monitor and determine whether the human gait is normal and further warn of falls. This type of solution can better reflect the frequency, speed and other information of gait, but it is not effective in predicting falls. Because the pressure on the soles of the feet can only partially reflect certain fall behaviors, when the pressure on the soles of the feet does not change, the upper body of the human body can also show many signs of falls, such as recoil falls, drooping falls, etc. Since the sensor is on the soles of the feet, it is inconvenient to wear, and the electronic devices can easily affect the comfort of the feet during walking, so it is not suitable for universal use.

[0009] 3) Existing fall warning systems often use remote transmission and calculation to feedback results. Wear relevant sensors for fall warning, and send the raw data of the worn sensors to the PC through the wireless transmission module, and finally process it through the algorithm on the PC. This type of solution requires a large amount of equipment and has a high algorithm complexity, which is not conducive to edge node calculation.

[0010] 4) Traditional fall warning algorithms based on acceleration signals and attitude angles have a high false alarm rate and often make fall judgments by setting acceleration signal thresholds and attitude angle thresholds. Although such solutions can filter out some behaviors similar to falls, they have no ability to recognize some actions in life and have specific requirements for the wearing direction of the sensor. Summary of the invention

[0011] The purpose of the present invention is to overcome the defects of the prior art and provide a fall and similar action warning method based on inertial sensing. The method comprises the following steps:

[0012] Obtain the three-axis acceleration signal and three-axis angular velocity signal of the target in the body coordinate system, and calculate the combined acceleration;

[0013] According to the mapping relationship between the body coordinate system and the world coordinate system, the three-axis acceleration signal and the three-axis angular velocity signal of the world coordinate system are obtained;

[0014] Calculate the three-axis motion acceleration in the world coordinate system;

[0015] By comparing the magnitude of the combined acceleration with a set threshold, determining whether the target is in a weightless state;

[0016] When it is determined that the target is in a weightless state, a fall warning is issued based on the instantaneous displacement in the vertical direction and / or the change in the inclination angle, wherein the instantaneous displacement is obtained based on the movement speed and movement acceleration of the target in the vertical direction, and the change in the inclination angle is obtained based on the angular velocity signals of the X-axis and the Y-axis in the world coordinate system.

[0017] Compared with the prior art, the advantages of the present invention are that, in order to solve the problems of high missed reporting rate and low accuracy of fall warning, by analyzing a large number of fall behaviors, typical features of fall behaviors are found and extracted, and through real-time analysis and processing, fall warning signals can be accurately predicted and issued before the fall, and the actuators can be notified to carry out timely fall protection and rescue. The present invention can filter out a large number of actions similar to fall behaviors in life and warn of common fall actions and special fall actions.

[0018] Further features and advantages of the present invention will become apparent from the following detailed description of exemplary embodiments of the present invention with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0020] Figure 1 is a flow chart of a method for early warning of falls and similar actions based on inertial sensing according to an embodiment of the present invention;

[0021] Figure 2 1 is a process diagram of a method for early warning of falls and similar actions based on inertial sensing according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless otherwise specifically stated.

[0023] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0024] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0025] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0026] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0027] In general, the provided inertial sensing-based fall and similar action warning method first collects the acceleration signal and angular velocity signal of human movement through an inertial posture sensor (such as MPU9250) (assuming that the sensor worn by the human body fits the human body very well, then the acceleration signal and angular velocity signal output by the sensor can be understood as the acceleration signal and angular velocity signal of human movement), and filters the original acceleration and angular velocity signals of the three axes (x-axis, y-axis, and z-axis) of the sensor.

[0028] Then, the combined acceleration value is calculated using the filtered three-axis acceleration values. The combined acceleration can reflect the weightlessness of the human body and can be used to determine whether the human body falls (falling is accompanied by weightlessness).

[0029] Next, the acceleration values ​​and angular velocity values ​​of the three axes of the sensor are mapped to the world coordinate system respectively. At this point, there are six quantities in the world coordinate system, namely the three-axis acceleration values ​​and three-axis angular velocity values ​​mapped by the sensor. The three-axis angular velocity values ​​in the world coordinate system represent the angular velocity values ​​of the human body during the tilt process around the geographical east / west direction, south / north direction, and vertical direction. For example, if the X-axis of the world coordinate system represents the east / west direction, when the human body falls toward the south / north direction, the angular velocity of the X-axis is the angular velocity of the human body during the tilt process around the east / west direction. Integrating it within a specific time interval will give the change in the tilt angle of the human body during the tilt process around the east / west direction. This idea will be used later to calculate the change in the tilt angle of the human body within a certain time period, and then to assist in determining whether a fall is about to occur.

[0030] Furthermore, considering that the inertial sensor will be affected by gravity, the value of the three-axis acceleration in the body coordinate system will contain the component of gravity acceleration (the magnitude and direction of the components of gravity acceleration on the three coordinate axes in the body coordinate system are determined by the sensor's own posture). Therefore, when the three-axis acceleration value of the sensor in the body coordinate system is mapped to the world coordinate system, there will be a vertically downward gravity acceleration value in the Z-axis direction of the world coordinate system. Then, the acceleration in the Z-axis direction in the world coordinate system is subtracted from the gravity acceleration to obtain the motion acceleration in the Z-axis direction. The accelerations on the remaining X-axis and Y-axis in the world coordinate system are also motion acceleration measurements, but the directions are perpendicular to the motion acceleration on the Z-axis and do not include gravity acceleration. There is no need to subtract the natural gravity acceleration, because the natural gravity acceleration is always vertically downward and only exists on the Z-axis.

[0031] Then, the value of the combined acceleration is judged, and the combined acceleration is used to filter out false falling actions and identify true falling actions. If it is less than the normal threshold, it indicates that the human body is in a state of weightlessness, that is, it begins to fall (weightlessness is an inherent attribute of falling behavior); otherwise, it indicates that the human body has no tendency to fall, and is in a fall interference action (for example, daily behavior such as walking and jogging) or a static state (the threshold used here is not the traditional threshold method, the threshold here is just a simple prediction, and the real fall warning method is detailed in the following steps). Then, the human body will be displaced in the vertical direction of the world coordinate system, and the acceleration of the displacement is the acceleration in the Z-axis direction, and the acceleration in the Z-axis direction is vertically downward. At the same time, the calculation of the instantaneous vertical displacement of the human body and the calculation of the instantaneous tilt angle change of the human body are triggered. Whether the human body has a tendency to fall, the key point is to monitor whether the center of gravity of the human body or a part of the human body other than the limbs continues to move downward and continues to have a large tilt change, so the displacement of the human body in the vertical direction and the change in the tilt angle of the human body are calculated.

[0032] For example, if the vertical displacement of the human body in the world coordinate system is greater than a certain threshold, and the change in the human body's inclination angle is greater than a certain threshold, a fall warning will be issued. The vertical displacement can be calculated by speed and time; the change in the inclination angle can be calculated by integrating the composite angular velocity of the X-axis and the Y-axis in the world coordinate system. In the calculation process of the above displacement and inclination angle change, once the monitoring of the human body's inclination angle change and the vertical downward displacement no longer continue to grow positively / negatively, the calculation of the above displacement and inclination angle will be stopped immediately and the calculated results will be cleared to prepare for the next fall.

[0033] For the sake of clarity, we first introduce the parameters involved, which are defined as follows: The three-axis acceleration and angular velocity of the inertial sensor (IMU) are a and x 、a y 、a z and ω x ,ω y ,ω z ; The three-axis Euler angles (attitude angles) are γ, ψ, and θ respectively; the total acceleration is a 合 ; Acceleration threshold a 阈值 is a constant; the three-axis acceleration, gravitational acceleration and angular velocity in the world coordinate system are a X 、a Y 、a Z and g and ω X ,ω Y ,ω Z , where g is the actual local gravity acceleration; the vertical speed of the human body is v 动 ; The acceleration of the human body in the vertical direction is Acc Z; The change in the human body tilt angle is Δφ, and the threshold value of the change in the human body tilt angle is Δφ 阈值 ; The human body weightlessness mark is G 失 ;IMU sampling period d t =1 / (sampling frequency); the vertical displacement of the human body due to weightlessness is s z ; The fall warning signal sign is Sign.

[0034] Specifically, combined Figure 1 and Figure 2 As shown, the provided fall and similar action warning method based on inertial sensing includes the following steps:

[0035] Step S110, using an inertial attitude sensor to collect acceleration signals, angular velocity signals and attitude angles of human body motion.

[0036] First, read the three-axis acceleration and attitude angle of the inertial attitude sensor. x = IMU x-axis acceleration input; a y = IMU y-axis acceleration input; a z = z-axis acceleration input of IMU; ω x = IMU's x-axis angular velocity input, ω y = IMU's y-axis angular velocity input, ω z =IMU's z-axis angular velocity input; γ =y-axis attitude angle; ψ =z-axis attitude angle; θ =x-axis attitude angle.

[0037] It should be understood that the attitude angle can be calculated in a variety of ways, such as a quaternion attitude solution algorithm based on an extended Kalman filter, a complementary filter method, or a gradient descent method.

[0038] Step S120, filtering the collected original signal to obtain a three-axis acceleration signal and a three-axis angular velocity signal in the body coordinate system.

[0039] For example, the collected original three-axis acceleration signal and three-axis angular velocity signal are filtered using a moving average filter to remove noise burrs. The general formula of the moving average filter is:

[0040]

[0041] Where windowsize is the set window size. The moving average filter moves a window of length windowsize along the data and calculates the average of the data contained in each window.

[0042] Specifically, the collected three-axis acceleration signal and three-axis angular velocity signal are filtered, that is:

[0043]

[0044] in, and It is an intermediate variable generated in the filtering process and has no actual meaning. t represents the time when the signal is collected.

[0045] After the above processing, the three-axis acceleration signal and the three-axis angular velocity signal in the body coordinate system can be obtained after filtering. It should be noted that in addition to using sliding window filtering, other filtering methods can also be used.

[0046] Step S130, calculating the three-axis acceleration signal and the three-axis angular velocity signal in the world coordinate system, and the motion acceleration in the vertical direction.

[0047] The posture angles of the human body obtained by the inertial sensor include the pitch angle γ, the roll angle θ and the heading angle ψ. The rotation matrix from the body coordinate system to the world coordinate system is: but equal:

[0048]

[0049] The three-axis acceleration value of the object in the world coordinate system is a X 、a Y 、a Z , the three-axis angular velocity value is ω X ,ω Y ,ω Z .

[0050] Among them, a X 、a Y 、a Z with a x 、a y 、a z The conversion relationship is as follows:

[0051]

[0052] Among them, ω X ,ω Y ,ω Z With ω x ,ω y ,ω z The conversion relationship is as follows:

[0053]

[0054] Among them, γ is the rotation angle on the y-axis in the body coordinate system, that is, the pitch angle, ψ is the rotation angle on the z-axis in the body coordinate system, that is, the yaw angle, and θ is the rotation angle on the x-axis in the body coordinate system, that is, the roll angle.

[0055] Furthermore, the interference of gravity acceleration is filtered out in the world coordinate system to obtain the motion acceleration in the Z-axis direction, then the three-axis motion acceleration Acc in the world coordinate system is X 、Acc Y 、Acc Z The calculation of is as follows:

[0056]

[0057] Among them, Acc Z is the acceleration of the Z axis in the world coordinate system, and g is the actual local gravity acceleration. Y is the acceleration of the motion in the Y-axis direction in the world coordinate system, Acc X It is the acceleration of motion in the X-axis direction in the world coordinate system.

[0058] In this step, according to the mapping relationship between the body coordinate system and the world coordinate system, the three-axis attitude angle is used to solve the mapping matrix, and the three-axis acceleration signal and angular velocity signal are mapped to the world coordinate system by the mapping matrix. Then, the three-axis acceleration signal in the world coordinate system is subtracted from the gravity acceleration g in the vertical Z axis to obtain the three-axis motion acceleration in the world coordinate system.

[0059] It should be noted that when calculating the net motion acceleration in the world coordinate system, the above embodiment first maps the three-axis acceleration value in the body coordinate system to the world coordinate system, and then subtracts the gravity acceleration. This process can also be replaced by the following steps: mapping the gravity acceleration in the world coordinate system to the body coordinate system, then subtracting the gravity acceleration in the body coordinate system from the three-axis acceleration in the body coordinate system, and then mapping the three-axis acceleration value minus the gravity acceleration to the world coordinate system. Similarly, the three-axis value in the world coordinate system is also the net motion acceleration.

[0060] Step S140, calculating the total acceleration in the body coordinate system.

[0061] For example, the total acceleration in the body coordinate system is expressed as:

[0062]

[0063] Among them, the three-axis acceleration value after filtering is a x 、a y 、a zThe three-axis angular velocity value after filtering is ω x ,ω y ,ω z The combined acceleration can reflect whether the human body is weightless and the degree of weightlessness.

[0064] Step S150, judging whether the target is in a weightless state based on the combined acceleration, and issuing a fall warning according to the instantaneous displacement in the vertical direction and the change in the tilt angle.

[0065] First, determine the weightlessness of the human body. If the combined acceleration a 合 <a 阈值 , it indicates that the object begins to be in a state of weightlessness. At the same time, the vertical velocity v of the human body is calculated. 动 and motion acceleration Acc Z And integrate the three-axis angular velocity in the world coordinate system. Then solve the vertical displacement s of the human body z and the change in the inclination angle relative to the vertical direction Δφ. For example, if the weightless displacement s z >s 阈值 And Δφ>Δφ 阈值 , it is judged that the human body is about to fall, and a fall warning signal is issued.

[0066] Specifically, if the combined acceleration is less than the normal threshold, that is: a 合 <a 阈值 , it means that the human body is weightless, so the weightlessness mark position is 1, that is, G 失 =1; otherwise, the weightlessness mark is set to 0, i.e. G 失 =0;

[0067] If G 失 =1, then start to calculate the instantaneous displacement in the vertical direction, and we have:

[0068]

[0069] Wherein, T represents the end time of calculating displacement S (that is, the time of issuing fall warning, that is, the time when Sign=1 in the following text), t represents the time index of calculating displacement, and t=1 represents the start time of calculating displacement S (that is, G 失 =1), d t Indicates the sampling period.

[0070] At the same time, the change in the tilt angle is calculated:

[0071]

[0072]

[0073] It should be noted that the body's own rotation around the Z axis has nothing to do with the falling process and is therefore not included in the calculation.

[0074] When the instantaneous displacement in the vertical direction is greater than the threshold value (i.e., s z >s 阈值 ), it means that the center of gravity or key parts of the human body are moving downward in a large range; when the change in the human body's inclination angle is greater than the threshold (i.e. Δφ>Δφ 阈值 ), it means that the center of gravity or key parts of the human body have tilted significantly. In one embodiment, when the above two variables are both greater than their corresponding thresholds, it is judged that the human body is about to fall, and then the fall warning sign is set to 1, that is, Sign = 1; if the judgment condition is not met, the fall warning sign is still set to 0, that is, Sign = 0.

[0075] If the combined acceleration is greater than the normal threshold, that is: s 合 >s 阈值 , then record s Z and the value of Δφ and and The values ​​of and Δφ T as well as and express.

[0076] In one embodiment, when s Z When and Δφ satisfy the following conditions 1 and 2 simultaneously or when ω x =0 or ω y = 0, then s Z The sum of Δφ values ​​is cleared to zero, ending the warning process for this fall, and returning to step S110 to prepare for the next fall warning; otherwise, jumping to determine whether the combined acceleration is less than the normal threshold to continue calculating the instantaneous displacement s in the vertical direction. Z And the change in the human body's inclination angle Δφ.

[0077] Condition one:

[0078] Condition two:

[0079] In this step S150, considering that normal human behavior involves short-term slight weightlessness, a threshold is set to distinguish between false weightlessness of normal behavior and true weightlessness that may cause falls. When in the true weightlessness state, the vertical displacement of the human body during the true weightlessness state is calculated, and the change in the tilt angle of the human body in the world coordinate system during the true weightlessness state is calculated. In addition, in a sampling period d of the inertial sensor tIn the real weightlessness state, when the displacement in the vertical direction is greater than the set displacement threshold and the change in the inclination angle in the world coordinate system is greater than the set threshold, it is considered that a fall is about to occur, triggering a fall warning signal. Through these designs, a large number of actions similar to fall behaviors in life (such as sitting, walking, standing, bending, lying down, squatting, jumping, etc.) can be filtered out and common fall actions (such as left / right fall, front / back fall, lateral fall, etc.) and special fall actions (such as falling down, falling from bed, falling during limping, falling on the hip / knee first, sliding slowly against the wall, etc.) can be warned.

[0080] In order to further verify the effect of the present invention, actions similar to falling behavior are verified. The experimental results are shown in Table 1. It can be seen that the average false alarm rate is only 1.40%.

[0081] Table 1 Experimental results

[0082] ADL action types Number of experiments Number of false positives False Positive Feedback Method False Positive Rate walk 150 1 Language Tips 0.7％ jogging 150 1 Language Tips 0.7％ Sit down and stand up 150 0 Language Tips 0.0％ Lie down 150 2 Language Tips 1.3％ Squat and stand up 150 3 Language Tips 2.0％ jump 150 2 Language Tips 1.3％ Go up the stairs 150 3 Language Tips 2.0％ Go down the stairs 150 2 Language Tips 1.3％ Intermittent falling movements 150 5 Language Tips 3.3％

[0083] It should be noted that, without violating the spirit of the present invention, those skilled in the art may make appropriate changes or modifications to the above embodiments. For example, the vertical displacement of the human body under true weightlessness or the change in the deviation angle of the human body relative to the vertical direction is used as a fall discrimination feature for fall warning; the body coordinate system is mapped to the world coordinate system and the gravity acceleration is filtered out to obtain the motion acceleration. For example, in other different scenarios, the combined acceleration is used as an important marker for judging the weightlessness of the human body to calculate the vertical position distance, or the combined acceleration is used as an important marker for judging the weightlessness of the human body to calculate the deviation angle relative to the vertical direction. Alternatively, the X-axis angular velocity and the Y-axis angular velocity in the world coordinate system are used, and the square and then the sum are taken, and finally the square root operation relationship is obtained to obtain the combined angular velocity used for fall warning or detection algorithm, and the method and idea of ​​obtaining the combined angular velocity are within the scope of protection. In addition, methods for determining whether fall monitoring is terminated by using the change in the combined acceleration and the vertical displacement in the vertical direction, or the combined acceleration and the offset angle, or the combined acceleration and the vertical displacement and the offset angle in the vertical direction are all within the scope of protection of the present invention.

[0084] In summary, the present invention first filters the original signal to remove interference and signal noise to ensure that subsequent calculations are accurate and not interfered by measurement noise; the entire fall warning process does not rely on the posture angle, and can prevent the crosstalk of daily behaviors similar to falls in life. More importantly, the present invention utilizes many common features of fall behaviors, namely: the displacement of the human body in the vertical direction in the world coordinate system in a weightless state and the change in the vertical offset angle both vary greatly. By monitoring the instantaneous displacement of the human body in the vertical direction and the change in the offset angle in the weightless state, the fall behavior is predicted, which improves the accuracy of the prediction. In the case of intermittent actions of fall behavior (such as slowly sliding against the wall), the present invention can effectively identify such fall behaviors by monitoring the change in the magnitude of the acceleration change, the change in the magnitude of the vertical displacement, and the direction of the angular velocity change, so that the distinction between fall behavior and non-fall behavior is higher. In addition, it has been verified by experiments that the present invention has low complexity, high execution efficiency, small computing power required, and can also respond and process in real time at the edge terminal, and has a wider application.

[0085] The present invention may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present invention.

[0086] Computer readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device. Computer readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (non-exhaustive list) of computer readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, for example, a punch card or a convex structure in a groove on which instructions are stored, and any suitable combination thereof. The computer readable storage medium used here is not interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated by a waveguide or other transmission medium (for example, a light pulse by an optical fiber cable), or an electrical signal transmitted by a wire.

[0087] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0088] The computer program instructions for performing the operation of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, Python, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed entirely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be personalized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present invention.

[0089] Various aspects of the present invention are described herein with reference to the flow charts and / or block diagrams of the methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each box of the flow chart and / or block diagram and the combination of each box in the flow chart and / or block diagram can be implemented by computer-readable program instructions.

[0090] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0091] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0092] The flow charts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a part of a module, a program segment or an instruction, and a part of the module, a program segment or an instruction contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented by a dedicated hardware-based system that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. It is well known to those skilled in the art that it is equivalent to implement it by hardware, implement it by software, and implement it by combining software and hardware.

[0093] Embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the marketplace, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein. The scope of the present invention is defined by the appended claims.

Claims

1. A fall and similar action warning method based on inertial sensing, comprising the following steps: Obtain the three-axis acceleration signal and three-axis angular velocity signal of the target in the body coordinate system, and calculate the combined acceleration; According to the mapping relationship between the body coordinate system and the world coordinate system, the three-axis acceleration signal and the three-axis angular velocity signal of the world coordinate system are obtained; Calculate the three-axis motion acceleration in the world coordinate system; By comparing the combined acceleration with a set threshold, determining whether the target is in a weightless state; When it is determined that the target is in a weightless state, a fall warning is performed according to the instantaneous displacement in the vertical direction and the change in the tilt angle, wherein the instantaneous displacement is obtained based on the movement speed and movement acceleration of the target in the vertical direction, and the change in the tilt angle is obtained based on the angular velocity signals of the X-axis and the Y-axis in the world coordinate system; Among them, a fall warning is performed when the following conditions are met: when the instantaneous displacement in the vertical direction is greater than a first threshold and when the change in the human body's inclination angle is greater than a second threshold; Wherein, the instantaneous displacement in the vertical direction is expressed as: The variation of the tilt angle is expressed as: The vertical velocity is v. 动 ; The vertical acceleration is Acc Z ; The change in the tilt angle is Δφ, and the threshold of the change in the tilt angle is Δφ 阈值 ;d t is the sampling period; the instantaneous displacement in the vertical direction is s Z , t represents the time index of displacement calculation, t=1 represents the starting time of displacement calculation, and T represents the ending time of displacement calculation; Among them, it also includes: When the target is judged to be in a non-weightless state, execute: Records Z , Δφ, and The values ​​of Δφ T , and express; when and and Will s Z The Δφ value is reset to zero, ending the early warning process for this fall; Among them, ω x ,ω y ,ω z is the three-axis angular velocity value in the body coordinate system, Δφ is the change in the tilt angle, s Z is the instantaneous displacement in the vertical direction, and T represents the end time of calculating the displacement.

2. The method according to claim 1, characterized in that The mapping relationship between the body coordinate system and the world coordinate system is expressed as: Among them, the three-axis acceleration signal in the body coordinate system is a x 、a y 、a z , the three-axis angular velocity signal is ω x ,ω y ,ω z , γ is the rotation angle on the y-axis in the body coordinate system, ψ is the rotation angle on the z-axis in the body coordinate system, and θ is the rotation angle on the x-axis in the body coordinate system. It is the rotation matrix from the body coordinate system to the world coordinate system.

3. The method according to claim 1, characterized in that The three-axis motion acceleration in the world coordinate system is expressed as: Among them, Acc Z is the acceleration of the Z axis in the world coordinate system, Acc Y is the acceleration of the motion in the Y-axis direction in the world coordinate system, Acc X is the acceleration of motion in the X-axis direction in the world coordinate system, and g is the acceleration due to gravity.

4. The method according to claim 1, characterized in that The three-axis acceleration signal and the three-axis angular velocity signal in the body coordinate system are obtained by filtering the original data with a moving average filter.

5. The method according to claim 1, characterized in that The resultant acceleration is expressed as: Among them, a x 、a y 、a z is the three-axis acceleration signal in the body coordinate system, a x Corresponding to the x-axis, a y Corresponding to the y-axis, a z Corresponding to the z-axis.

6. A computer-readable storage medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

7. A computer device comprising a memory and a processor, wherein a computer program capable of running on the processor is stored in the memory, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

Images