Disclosure of Invention
The invention mainly aims to provide a fall detection early warning method, a fall detection early warning device, fall detection equipment and a storage medium, and aims to solve the technical problem of how to simply and effectively detect falling of a person to be detected.
In order to achieve the above object, the present invention provides a fall detection early warning method, which includes the following steps:
Performing balance detection on a person to be detected according to gesture information corresponding to the person to be detected to obtain a balance detection result;
Determining unbalance time length of the person to be detected in an unbalance state according to the balance detection result;
And when the unbalance time length is longer than the preset time length, carrying out falling detection on the personnel to be detected, and carrying out early warning according to a falling detection result.
Optionally, the step of performing balance detection on the person to be detected according to the gesture information corresponding to the person to be detected to obtain a balance detection result specifically includes:
Collecting a plurality of point cloud data corresponding to the personnel to be detected through a radar;
acquiring a motion signal corresponding to a person to be detected through a motion sensor, and judging the accuracy of the plurality of point cloud data according to the motion signal;
When the accuracy judgment is passed, acquiring a thermal image corresponding to the person to be detected through an infrared sensor array, and determining posture change information according to the plurality of point cloud data and the thermal image;
and carrying out balance detection on the personnel to be detected according to the posture change information to obtain a balance detection result.
Optionally, the step of collecting, by a motion sensor, a motion signal corresponding to a person to be detected, and performing accuracy judgment on the plurality of point cloud data according to the motion signal specifically includes:
Acquiring a motion signal corresponding to a person to be detected through a motion sensor, and preprocessing the motion signal to obtain a processed signal;
determining the position information of the motion sensor at each moment according to the processed signals, and acquiring the wearing position corresponding to the motion sensor;
Selecting target point cloud data corresponding to the wearing position from the plurality of point cloud data;
and accurately judging the plurality of point cloud data according to the position information and the target point cloud data.
Optionally, the step of performing balance detection on the person to be detected according to the posture change information to obtain a balance detection result specifically includes:
Collecting pressure data through a flexible pressure sensor arranged at a preset position corresponding to the person to be detected;
according to the attitude change information and the pressure data, determining gravity center change information corresponding to the person to be detected, and determining gravity center change speed corresponding to the gravity center change information;
Determining a moving track corresponding to the person to be detected according to the gesture change information, and carrying out anomaly detection on the moving track to obtain an anomaly detection result;
And when the abnormality detection result is that the movement track is abnormal and the gravity center change speed is greater than a preset speed, determining that the balance detection result is that the person to be detected is in an unbalanced state.
Optionally, when the unbalance time is longer than a preset time, performing fall detection on the person to be detected, and performing early warning according to a fall detection result, including:
When the unbalance time length is longer than a preset time length, acquiring an initial highest distance between the person to be detected and the ground at the current time and an audio signal generated by the person to be detected;
judging whether the initial highest distance is accurate or not according to a detection scene of the person to be detected;
When the initial highest distance is inaccurate, determining a target highest distance according to the detection scene;
Identifying the audio signal through a preset voice identification algorithm to obtain keywords;
and when the highest distance of the target is smaller than a preset distance and the preset word exists in the keyword, determining that the falling detection result is that the personnel to be detected is in a falling state, and carrying out early warning.
Optionally, when the unbalance time is longer than the preset time, the step of obtaining the initial highest distance between the person to be detected and the ground at the current time specifically includes:
When the unbalance time length is longer than the preset time length, acquiring current point cloud data corresponding to the person to be detected at the current time through a radar;
acquiring current point cloud coordinates corresponding to the current point cloud data, and determining a current point cloud distance set according to the current point cloud coordinates;
And taking the maximum value in the current point cloud distance set as the initial highest distance between the person to be detected and the ground.
Optionally, the step of determining whether the initial highest distance is accurate according to the detection scene where the person to be detected is located specifically includes:
determining interference point cloud data corresponding to a detection scene where the person to be detected is located, and determining an interference distance corresponding to the interference point cloud data;
If the difference value between the interference distance and the initial highest distance is smaller than a preset threshold value, judging that the initial highest distance is inaccurate;
correspondingly, when the initial highest distance is inaccurate, determining the highest distance of the target according to the detection scene specifically includes:
When the initial highest distance is inaccurate, screening the current point cloud distance set based on the interference distance to obtain a screened point cloud distance set;
and taking the maximum value in the screened point cloud distance set as the target highest distance.
In addition, in order to achieve the above object, the present invention also provides an early warning device for fall detection, the early warning device for fall detection including:
the balance detection module is used for carrying out balance detection on the personnel to be detected according to the gesture information corresponding to the personnel to be detected, so as to obtain a balance detection result;
The time length determining module is used for determining the unbalance time length of the person to be detected in the unbalance state according to the balance detection result;
and the falling detection module is used for carrying out falling detection on the personnel to be detected when the unbalance time is longer than the preset time and carrying out early warning according to a falling detection result.
In addition, in order to achieve the above object, the present invention also proposes an early warning device for fall detection, the early warning device for fall detection comprising: the fall detection early warning device comprises a memory, a processor and a fall detection early warning program stored on the memory and capable of running on the processor, wherein the fall detection early warning program is configured to realize the steps of the fall detection early warning method.
In addition, in order to achieve the above object, the present invention further proposes a storage medium having stored thereon a fall detection pre-warning program, which when executed by a processor, implements the steps of the fall detection pre-warning method as described above.
According to the invention, balance detection is carried out on the personnel to be detected according to the posture information corresponding to the personnel to be detected, a balance detection result is obtained, then the unbalance time length of the personnel to be detected in an unbalance state is determined according to the balance detection result, when the unbalance time length is longer than the preset time length, the personnel to be detected is subjected to fall detection, and early warning is carried out according to the fall detection result. According to the invention, balance detection is carried out on the personnel to be detected according to the posture information corresponding to the personnel to be detected, the unbalance time length of the personnel to be detected in an unbalance state is determined according to the balance detection result, when the unbalance time length is longer than the preset time length, the personnel to be detected possibly has a falling risk, and then the personnel to be detected is subjected to falling detection, so that the falling detection can be avoided when the personnel to be detected is in a short unbalance state, the falling detection accuracy can be improved, and the falling detection can be simply and effectively carried out on the personnel to be detected.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, fig. 1 is a schematic structural diagram of an early warning device for fall detection in a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the fall detection early warning device may include: a processor 1001, such as a central processing unit (Central Processing Unit, CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a Wireless interface (e.g., a Wireless-Fidelity (Wi-Fi) interface). The Memory 1005 may be a high-speed random access Memory (Random Access Memory, RAM) or a stable nonvolatile Memory (NVM), such as a disk Memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
It will be appreciated by those skilled in the art that the structure shown in fig. 1 does not constitute a limitation of the fall detection early warning device, and may include more or fewer components than shown, or certain components may be combined, or a different arrangement of components.
As shown in fig. 1, an operating system, a network communication module, a user interface module, and a fall detection early warning program may be included in the memory 1005 as one storage medium.
In the fall detection early warning device shown in fig. 1, the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with a user; the processor 1001 and the memory 1005 in the early warning device for fall detection of the present invention may be disposed in the early warning device for fall detection, where the early warning device for fall detection invokes the early warning program for fall detection stored in the memory 1005 through the processor 1001, and executes the early warning method for fall detection provided by the embodiment of the present invention.
Based on the foregoing fall detection early warning device, an embodiment of the present invention provides a fall detection early warning method, and referring to fig. 2, fig. 2 is a schematic flow chart of a first embodiment of the fall detection early warning method of the present invention.
In this embodiment, the early warning method for fall detection includes the following steps:
step S10: and carrying out balance detection on the personnel to be detected according to the gesture information corresponding to the personnel to be detected, and obtaining a balance detection result.
It should be noted that, the execution body of the embodiment may be a computing service device with functions of data processing, network communication and program running, such as a computer, or an electronic device or a fall detection early warning device capable of implementing the above functions. The present embodiment and the following embodiments will be described below by taking the fall detection warning device as an example.
It is understood that the person to be detected refers to a person who needs to perform fall detection, and may be a person who needs to perform fall detection, such as an elderly person, a child, or the like. Posture information refers to information related to the posture of a human body, such as standing, lying, sitting, etc.
In a specific implementation, the balance detection can be performed on the person to be detected according to the gesture information to obtain a balance detection result, and whether the gesture information changes or not can be judged in a preset time period, if not, the balance detection result is that the person to be detected is in a balance state.
Step S20: and determining the unbalance duration of the person to be detected in the unbalance state according to the balance detection result.
It can be understood that the unbalance time length refers to a time length when a person to be detected is in an unbalance state, specifically, a starting time when a balance detection result is that the person to be detected is in the unbalance state can be obtained first, a time when the person to be detected is changed from the unbalance state to the balance state is taken as an ending time, a time length between the ending time and the starting time is an unbalance time length, and if the person to be detected is always in the unbalance state, the unbalance time length is a time length between the current time and the starting time.
Step S30: and when the unbalance time length is longer than the preset time length, carrying out falling detection on the personnel to be detected, and carrying out early warning according to a falling detection result.
It should be understood that the preset duration is the longest duration that the preset person to be detected can be in the unbalanced state, and the preset durations may be different for the person to be detected in different age groups, and the preset durations may be 1 second, 1.5 seconds, etc., which is not particularly limited in this embodiment.
In a specific implementation, when the unbalance time length is longer than the preset time length, the time that the personnel to be detected is in the unbalance state is excessively long, the personnel to be detected has a falling risk, at the moment, the personnel to be detected can fall to be detected to obtain a falling detection result, when the falling detection result is that the personnel to be detected is in the falling state, early warning is needed, and the early warning mode can be that a short message is sent to an emergency contact person of the personnel to be detected.
According to the method, balance detection is carried out on the person to be detected according to the posture information corresponding to the person to be detected, a balance detection result is obtained, then the unbalance duration of the person to be detected in an unbalance state is determined according to the balance detection result, when the unbalance duration is longer than the preset duration, falling detection is carried out on the person to be detected, and early warning is carried out according to the falling detection result. According to the method, balance detection is firstly carried out on the personnel to be detected according to the gesture information corresponding to the personnel to be detected, the unbalance duration of the personnel to be detected in an unbalance state is determined according to the balance detection result, when the unbalance duration is longer than the preset duration, the personnel to be detected possibly have a falling risk, then the personnel to be detected are subjected to falling detection, the falling detection can be avoided when the personnel to be detected are in a short-term unbalance state, and accordingly the falling detection accuracy can be improved, and the falling detection is simply and effectively carried out on the personnel to be detected.
Referring to fig. 3, fig. 3 is a flowchart illustrating a second embodiment of a fall detection early warning method according to the present invention.
Based on the first embodiment, in this embodiment, the step S10 includes:
step S101: and acquiring a plurality of point cloud data corresponding to the person to be detected through a radar.
It can be appreciated that, in this embodiment, a plurality of point cloud data corresponding to a person to be detected may be collected by a radar, specifically, may be a laser radar, a millimeter wave radar, or the like, and the plurality of point cloud data may be point cloud data corresponding to an edge contour of the person to be detected.
Step S102: and acquiring a motion signal corresponding to a person to be detected through a motion sensor, and judging the accuracy of the plurality of point cloud data according to the motion signal.
It should be understood that the motion sensor may be worn on the body of the person to be detected, may be worn on the wrist, the waist, or the like, and the number of the motion sensors is not particularly limited in this embodiment.
It is understood that the motion sensor may be a speed sensor, an acceleration sensor, or the like, and the motion signal collected is a speed signal when the motion sensor is a speed sensor, and an acceleration signal when the motion sensor is an acceleration sensor. And accurately judging the plurality of point cloud data according to the motion signals, namely judging whether the point cloud data are correct or not.
Further, in order to effectively determine the accuracy of the point cloud data, in this embodiment, the step S102 includes: acquiring a motion signal corresponding to a person to be detected through a motion sensor, and preprocessing the motion signal to obtain a processed signal; determining the position information of the motion sensor at each moment according to the processed signals, and acquiring the wearing position corresponding to the motion sensor; selecting target point cloud data corresponding to the wearing position from the plurality of point cloud data; and accurately judging the plurality of point cloud data according to the position information and the target point cloud data.
It should be understood that, in this embodiment, the motion signal collected by the motion sensor may be preprocessed, where the preprocessing may include filtering, noise reduction, amplification, and so on, to obtain a processed signal.
It can be understood that the position information of the motion sensor at each time point can be determined according to the processed signal, and when the processed signal is a speed signal, the position information of the motion sensor at each time point can be obtained based on the integration of the speed along with time, and the wearing position refers to the position of the motion sensor worn on the person to be detected. The target point cloud data corresponding to the wearing position may be selected from the plurality of point cloud data, for example, when the wearing position is a wrist, the target point cloud data at the wrist may be selected from the plurality of point cloud data.
In a specific implementation, accuracy judgment can be performed on the plurality of point cloud data according to the position information and the target point cloud data, specifically, the target point cloud data can include point cloud data under a plurality of moments, position change information corresponding to wearing positions can be obtained according to the plurality of target point cloud data, and because the target point cloud data can include three-dimensional coordinate data, the position change information can be obtained according to the three-dimensional coordinate data, then the position change information corresponding to the target point cloud data is compared with the position information corresponding to the motion signals, and if the position change information is basically the same, the accuracy judgment of the plurality of point cloud data is proved to pass.
Step S103: and when the accuracy judgment is passed, acquiring a thermal image corresponding to the person to be detected through an infrared sensor array, and determining posture change information according to the plurality of point cloud data and the thermal image.
It can be understood that when the accuracy judgment of the plurality of point cloud data passes, the thermal image corresponding to the person to be detected can be acquired through the infrared sensor array, and the infrared sensor array can be arranged around the person to be detected, for example, can be arranged at positions of 3 meters, 4 meters and the like near the person to be detected. The thermal image can be filtered by an adaptive filter to obtain a processed thermal image.
It should be understood that, in this embodiment, the posture change information may be determined according to the plurality of point cloud data and the processed thermal image, specifically, the edge profile in the processed thermal image may be obtained by an edge detection algorithm, and then the three-dimensional coordinate information of the edge profile of the person to be detected may be obtained according to the plurality of point cloud data obtained at each moment, that is, the edge profile of the person to be detected at each moment may be obtained according to the three-dimensional coordinate information, and the posture information at each moment may be obtained according to the edge profile, and finally the posture change information may be obtained, for example, the posture change information may be changed from lying down to lying on side, and vice versa.
In a specific implementation, when the accuracy judgment of the plurality of point cloud data is failed, the point cloud data can be collected through the radar again until the accuracy judgment is passed, and if the accuracy judgment is not passed for a long time, the radar can be replaced.
Step S104: and carrying out balance detection on the personnel to be detected according to the posture change information to obtain a balance detection result.
Further, in order to accurately perform balance detection on the person to be detected, in this embodiment, the step S104 includes: collecting pressure data through a flexible pressure sensor arranged at a preset position corresponding to the person to be detected; according to the attitude change information and the pressure data, determining gravity center change information corresponding to the person to be detected, and determining gravity center change speed corresponding to the gravity center change information; determining a moving track corresponding to the person to be detected according to the gesture change information, and carrying out anomaly detection on the moving track to obtain an anomaly detection result; and when the abnormality detection result is that the movement track is abnormal and the gravity center change speed is greater than a preset speed, determining that the balance detection result is that the person to be detected is in an unbalanced state.
It should be noted that, the preset position in this embodiment may be a sole, that is, the flexible pressure sensor may be disposed on the sole, and may collect a pressure signal output by the flexible pressure sensor and convert the pressure signal into pressure data, where the pressure data may represent a pressure change situation between the sole and the ground. The greater the pressure between the sole and the ground, the lower the centre of gravity of the person to be detected.
It should be understood that the gravity center change information refers to a change situation of the gravity center of a person to be detected, specifically, the gravity center change information can be determined according to gesture change information and pressure data, when the gesture information and the pressure data change, the corresponding gravity center can also change, a human body model can be constructed according to gesture information at each moment, then the human body model is analyzed, the gravity center information at each moment is obtained by combining the pressure data, the gravity center change information is formed, the gravity center change speed corresponding to the gravity center change information, namely, the gravity center change speed is determined, specifically, the average value of interval duration corresponding to the gravity center change information can be calculated, and then the average value is divided by the total duration of the obtained gravity center change information, so that the gravity center change speed is obtained.
It can be understood that the movement track of the person to be detected can also be determined according to the posture change information, specifically, the movement track can be obtained according to the change information of a plurality of body parts at each moment in combination with the posture change information, for example, the movement track can be obtained according to the change information of the head and the posture change information, if the head information suddenly descends from a certain height and the posture change information is unchanged, the movement track can be determined to be in an elevator at the moment, if the posture change information suddenly changes from standing to lying, and the movement track at the moment can be fallen down. And performing abnormality detection on the moving track, namely judging whether the moving track is normal, specifically, presetting a plurality of normal tracks, such as walking, elevator sitting and the like, if the moving track is not matched with the normal track, indicating that an abnormality detection result corresponding to the track is abnormal.
In a specific implementation, when the abnormality detection result is that the movement track is abnormal and when the gravity center change speed is greater than the preset speed, the balance detection result is determined that the person to be detected is in an unbalanced state, for example, when the movement track is a fall and the gravity center change speed in a time period corresponding to the movement track is greater than the preset speed, the person to be detected is determined to be in an unbalanced state.
According to the embodiment, the radar is used for collecting the plurality of point cloud data corresponding to the person to be detected, then the motion sensor is used for collecting the motion signal corresponding to the person to be detected, accuracy judgment is carried out on the plurality of point cloud data according to the motion signal, when the accuracy judgment passes, the infrared sensor array is used for collecting the thermal image corresponding to the person to be detected, gesture change information is determined according to the plurality of point cloud data and the thermal image, and balance detection is carried out on the person to be detected according to the gesture change information, so that a balance detection result is obtained. According to the method and the device, accuracy judgment is carried out on a plurality of point cloud data collected by the radar according to the motion signals collected by the motion sensor, whether the point cloud data are accurate or not can be effectively judged, when accuracy judgment is passed, balance detection is carried out on personnel to be detected according to the thermal image and attitude change information obtained by the point cloud data, and accuracy of the balance detection can be improved.
Referring to fig. 4, fig. 4 is a flowchart illustrating a third embodiment of a fall detection early warning method according to the present invention.
Based on the above embodiments, in this embodiment, the step S30 includes:
Step S301: when the unbalance time length is longer than the preset time length, acquiring the initial highest distance between the person to be detected and the ground at the current time and the audio signal generated by the person to be detected.
It can be understood that when the unbalance time is longer than the preset time, the initial highest distance between the person to be detected and the ground at the current time can be obtained, and the initial highest distance can be the highest distance between the person to be detected and the ground. The method can also acquire the audio signals generated by the personnel to be detected at the current moment, the audio signals can be all the audio signals which can be acquired in the environment where the personnel to be detected are located, the audio signals are subjected to noise reduction processing, and particularly the noise reduction processing can be performed through the audio noise reduction algorithms such as a filtering algorithm, a spectral subtraction algorithm and the like, so that noise-reduced signals are obtained, and more environmental noise, noise of other personnel and the like are avoided in the audio signals.
Further, in order to accurately obtain the initial highest distance, in this embodiment, the step S301 includes: when the unbalance time length is longer than the preset time length, acquiring current point cloud data corresponding to the person to be detected at the current time through a radar; acquiring current point cloud coordinates corresponding to the current point cloud data, and determining a current point cloud distance set according to the current point cloud coordinates; and taking the maximum value in the current point cloud distance set as the initial highest distance between the person to be detected and the ground.
It should be understood that when the unbalance time period is longer than the preset time period, the time period for indicating that the person to be detected is in the unbalance state is too long, and at this time, there may be a fall risk, and fall detection is required. The current point cloud data corresponding to the person to be detected at the current moment can be obtained through a radar, and the radar can be a laser radar, a millimeter wave radar and the like.
It can be understood that a current point cloud coordinate corresponding to the current point cloud data may be obtained, the current point cloud coordinate may be a three-dimensional coordinate, and a current point cloud distance set may be determined according to the current point cloud coordinate, where the current point cloud distance set may include a plurality of current point cloud distances, and the current point cloud distance may be z-axis data in the current point cloud coordinate. And taking the maximum value in the current point cloud set as an initial highest distance.
Step S302: judging whether the initial highest distance is accurate or not according to the detection scene of the person to be detected.
It should be understood that the detection scene may be a scene where a person to be detected is located, for example, an indoor, a factory, etc., and whether the initial highest distance is accurate is determined according to the detection scene.
Further, in order to effectively determine whether the initial maximum distance is accurate, in this embodiment, the step S302 includes: determining interference point cloud data corresponding to a detection scene where the person to be detected is located, and determining an interference distance corresponding to the interference point cloud data; and if the difference value between the interference distance and the initial highest distance is smaller than a preset threshold value, judging that the initial highest distance is inaccurate.
It can be understood that the interference point cloud data corresponding to the detection scene can be determined, for example, when the detection scene is a factory, the possible interference point cloud data is the point cloud data corresponding to dust or water drops, whether the point cloud data is the interference point cloud data can be judged according to the time length required for collecting the point cloud data, and if the time length is obviously inconsistent, the corresponding point cloud data can be used as the interference point cloud data.
It should be appreciated that the interference distance corresponding to the interference point cloud data may be determined, the interference distance may be a value of a z-axis in the coordinate data in the interference point cloud data, and then a difference between each interference distance and an initial highest distance is calculated, and if there is a difference less than a preset threshold, the initial highest distance is determined to be inaccurate, and the preset threshold is smaller, for example, 1 cm, 2 cm, etc. If all the differences are larger than the preset threshold value, the initial highest distance is judged to be accurate.
Step S303: and when the initial highest distance is inaccurate, determining the highest distance of the target according to the detection scene.
Further, in order to accurately obtain the target highest distance, the step S303 includes: when the initial highest distance is inaccurate, screening the current point cloud distance set based on the interference distance to obtain a screened point cloud distance set; and taking the maximum value in the screened point cloud distance set as the target highest distance.
It can be understood that when the initial highest distance is inaccurate, the current point cloud distance set can be screened based on the interference distance, that is, the point cloud distance, in which the difference value between the current point cloud distance set and the interference distance is smaller than the preset threshold value, is deleted, so as to obtain the screened point cloud distance set. And taking the maximum value in the screened point cloud distance set as the target highest distance.
Step S304: identifying the audio signal through a preset voice identification algorithm to obtain keywords;
It should be understood that the preset speech recognition algorithm is a technology for converting a speech signal into a text form, the basic principle of speech recognition includes preprocessing, feature extraction and modeling, the preprocessing process includes sampling, filtering and segmentation of the speech signal to convert the speech signal into a digital signal, the feature extraction is to convert the speech signal into a mathematical representation, and the modeling is to construct a speech model between the speech signal and the text. Common methods for feature extraction include mel-frequency cepstral coefficients and parametric model hidden markov models.
In a specific implementation, the embodiment may identify the audio signal by a preset voice recognition algorithm to obtain text information, and extract keywords from the text information, where the keywords may include exclamation, word of speech, verb, etc., and specifically may preset the type of the keywords.
Step S305: and when the highest distance of the target is smaller than a preset distance and the preset word exists in the keyword, determining that the falling detection result is that the personnel to be detected is in a falling state, and carrying out early warning.
It should be understood that when the highest target distance is smaller than the preset distance, the preset distance may be determined according to the stature of the person to be detected, specifically may be set as the maximum distance from the ground when the person to be detected lies on his side, and when the preset word exists in the keyword, the preset word may be "o", "careful", and the like, and at this time, it may be determined that the falling detection result is that the person to be detected is in a falling state, and early warning is performed.
When the unbalance time length is longer than the preset time length, the initial highest distance between the person to be detected and the ground at the current time and the audio signal generated by the person to be detected are obtained, whether the initial highest distance is accurate or not is judged according to the detection scene where the person to be detected is located, when the initial highest distance is inaccurate, the target highest distance is determined according to the detection scene, then the audio signal is identified through a preset voice recognition algorithm, keywords are obtained, and when the target highest distance is smaller than the preset distance and the preset words exist in the keywords, the falling detection result is determined that the person to be detected is in a falling state, and early warning is carried out. According to the method and the device for detecting the falling of the people to be detected, whether the initial highest distance is accurate is judged according to the detection scene where the people to be detected are located, when the initial highest distance is inaccurate, the target highest distance is determined according to the detection scene, accuracy of determining the highest distance between the people to be detected and the ground can be improved, and when the target highest distance is smaller than a preset distance and a preset word exists in a keyword, the falling detection result is determined to be that the people to be detected are in a falling state, so that accuracy of falling detection is improved.
Referring to fig. 5, fig. 5 is a block diagram illustrating a first embodiment of a fall detection warning device according to the present invention.
As shown in fig. 5, the early warning device for fall detection according to the embodiment of the present invention includes:
The balance detection module 10 is configured to perform balance detection on a person to be detected according to gesture information corresponding to the person to be detected, so as to obtain a balance detection result;
A time length determining module 20, configured to determine an unbalance time length of the person to be detected in an unbalance state according to the balance detection result;
The fall detection module 30 is configured to perform fall detection on the person to be detected when the unbalance time is longer than a preset time, and perform early warning according to a fall detection result.
According to the method, balance detection is carried out on the person to be detected according to the posture information corresponding to the person to be detected, a balance detection result is obtained, then the unbalance time length of the person to be detected in an unbalance state is determined according to the balance detection result, when the unbalance time length is longer than the preset time length, falling detection is carried out on the person to be detected, and early warning is carried out according to the falling detection result. According to the method, balance detection is firstly carried out on the personnel to be detected according to the gesture information corresponding to the personnel to be detected, the unbalance duration of the personnel to be detected in an unbalance state is determined according to the balance detection result, when the unbalance duration is longer than the preset duration, the personnel to be detected possibly have a falling risk, then the personnel to be detected are subjected to falling detection, the falling detection can be avoided when the personnel to be detected are in a short-term unbalance state, and accordingly the falling detection accuracy can be improved, and the falling detection is simply and effectively carried out on the personnel to be detected.
It should be noted that the above-described working procedure is merely illustrative, and does not limit the scope of the present invention, and in practical application, a person skilled in the art may select part or all of them according to actual needs to achieve the purpose of the embodiment, which is not limited herein.
In addition, technical details not described in detail in the present embodiment can be referred to the early warning method for fall detection provided in any embodiment of the present invention, and will not be described here again.
Based on the first embodiment of the early warning device for fall detection of the present invention, a second embodiment of the early warning device for fall detection of the present invention is provided.
In this embodiment, the balance detection module 10 is further configured to collect, by using a radar, a plurality of point cloud data corresponding to the person to be detected; acquiring a motion signal corresponding to a person to be detected through a motion sensor, and judging the accuracy of the plurality of point cloud data according to the motion signal; when the accuracy judgment is passed, acquiring a thermal image corresponding to the person to be detected through an infrared sensor array, and determining posture change information according to the plurality of point cloud data and the thermal image; and carrying out balance detection on the personnel to be detected according to the posture change information to obtain a balance detection result.
Further, the balance detection module 10 is further configured to collect a motion signal corresponding to a person to be detected through a motion sensor, and perform preprocessing on the motion signal to obtain a processed signal; determining the position information of the motion sensor at each moment according to the processed signals, and acquiring the wearing position corresponding to the motion sensor; selecting target point cloud data corresponding to the wearing position from the plurality of point cloud data; and accurately judging the plurality of point cloud data according to the position information and the target point cloud data.
Further, the balance detection module 10 is further configured to collect pressure data through a flexible pressure sensor disposed at a preset position corresponding to the person to be detected; according to the attitude change information and the pressure data, determining gravity center change information corresponding to the person to be detected, and determining gravity center change speed corresponding to the gravity center change information; determining a moving track corresponding to the person to be detected according to the gesture change information, and carrying out anomaly detection on the moving track to obtain an anomaly detection result; and when the abnormality detection result is that the movement track is abnormal and the gravity center change speed is greater than a preset speed, determining that the balance detection result is that the person to be detected is in an unbalanced state.
Further, the fall detection module 30 is further configured to obtain an initial highest distance between the person to be detected and the ground at the current time and an audio signal generated by the person to be detected when the unbalance time is longer than a preset time; judging whether the initial highest distance is accurate or not according to a detection scene of the person to be detected; when the initial highest distance is inaccurate, determining a target highest distance according to the detection scene; identifying the audio signal through a preset voice identification algorithm to obtain keywords; and when the highest distance of the target is smaller than a preset distance and the preset word exists in the keyword, determining that the falling detection result is that the personnel to be detected is in a falling state, and carrying out early warning.
Further, the fall detection module 30 is further configured to obtain, when the unbalance time period is longer than a preset time period, current point cloud data corresponding to the person to be detected at a current time period through a radar; acquiring current point cloud coordinates corresponding to the current point cloud data, and determining a current point cloud distance set according to the current point cloud coordinates; and taking the maximum value in the current point cloud distance set as the initial highest distance between the person to be detected and the ground.
Further, the fall detection module 30 is further configured to determine interference point cloud data corresponding to a detection scene where the person to be detected is located, and determine an interference distance corresponding to the interference point cloud data; if the difference value between the interference distance and the initial highest distance is smaller than a preset threshold value, judging that the initial highest distance is inaccurate; correspondingly, when the initial highest distance is inaccurate, determining the highest distance of the target according to the detection scene specifically includes: when the initial highest distance is inaccurate, screening the current point cloud distance set based on the interference distance to obtain a screened point cloud distance set; and taking the maximum value in the screened point cloud distance set as the target highest distance.
Other embodiments or specific implementation manners of the fall detection early warning device of the present invention may refer to the above method embodiments, and will not be described herein.
In addition, the embodiment of the invention also provides a storage medium, wherein the storage medium stores a fall detection early warning program, and the fall detection early warning program realizes the steps of the fall detection early warning method when being executed by a processor.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
From the above description of embodiments, it will be clear to a person skilled in the art that the above embodiment method may be implemented by means of software plus a necessary general hardware platform, but may of course also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. read-only memory/random-access memory, magnetic disk, optical disk), comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.