Household tumble detection system based on two-way conversation of Internet of Things

Abstract

The invention relates to the field of home tumble detection, aims to solve the problems that a home tumble detection system for old people is single in detection means, is easily interfered, affects the detection accuracy, and is difficult to carry out two-way communication at the first time after tumble, and particularly relates to a home tumble detection system based on two-way communication of the Internet of Things. According to the invention, the falling situation possibly existing in the home environment is collected through three modes of millimeter wave radar, sound detection and image collection, so that the accuracy of judging the falling situation is improved, the falling situation caused by missed judgment and misjudgment is avoided, the safety of the detected old people is ensured, and after the falling situation occurs, the safety of the old people is ensured. According to the system, the two-way communication function is automatically accessed through the system, so that air two-way communication is realized by means of the fixed sound acquisition assembly and the sound playing assembly, and when the elderly fall down, the elderly can be timely contacted with guardians or managers, so that the falling elderly can be timely rescued.

Description

Technical Field

[0001] The present invention relates to the field of home fall detection, specifically a home fall detection system based on two-way communication over the Internet of Things. Background Art

[0002] As people age, the consequences of falls become more severe. Compared with young people, the bones of the elderly are very fragile. Once the elderly fall, they are prone to hip fractures, lumbar fractures, etc. when landing on the buttocks, which will affect normal standing and walking. Once lying in bed for a long time, they are prone to problems such as infections, pressure sores, and deep vein thrombosis in the lower extremities. Therefore, how to promptly detect falls of the elderly at home is of great significance.

[0003] Currently, a technical solution is disclosed in the existing patent application CN115657004A. This solution preprocesses the point cloud data obtained from a millimeter-wave radar, normalizes the tracking target point cloud track, and then calls the corresponding model parameters. The pre-trained personnel fall detection model detects the personnel fall posture based on the normalized point cloud data, which can detect the home fall situation of the elderly to a certain extent. However, in this solution, there is only one means of obtaining falls, namely the millimeter-wave radar. Therefore, the judgment result depends on a single information collection means, resulting in the detection effect being easily interfered with, and thus it is impossible to efficiently and accurately detect the falls of the elderly. At the same time, after obtaining the fall information in this solution, there is a lack of a direct and effective integrated communication means. Therefore, communication still needs to be carried out through traditional communication devices such as mobile phones. After the elderly fall, if it is difficult to communicate through the mobile phone, it is not easy to conduct two-way communication, and thus it is impossible to directly understand the status of the fallen elderly.

[0004] In view of the above technical problems, the present application proposes a solution. Summary of the Invention

[0005] In the present invention, three methods, namely millimeter-wave radar, sound detection, and image acquisition, are used to simultaneously collect possible fall situations in the home environment, improving the accuracy of fall situation judgment, avoiding missed and misjudged fall situations, and ensuring the safety of the detected elderly. After a fall situation occurs, the system automatically accesses the two-way communication function, and thus, with the help of fixed sound collection components and sound playback components, two-way communication through the air is realized. When the elderly have a fall situation, they can be contacted in a timely manner with guardians or management personnel, so as to promptly rescue the fallen elderly, solve the problems of the home fall detection system having a single detection means, being easily interfered with, affecting the detection accuracy, and being difficult to conduct two-way communication immediately after a fall, and a home fall detection system based on two-way communication over the Internet of Things is proposed.

[0006] The object of the present invention can be achieved by the following technical solutions:

[0007] A home fall detection system based on two-way communication of the Internet of Things, comprising an identification information acquisition unit, a terminal discrimination unit, a two-way communication unit, a communication guarantee unit, and an alarm trigger unit. The identification information acquisition unit is used to collect fall judgment information and send the collected fall judgment information to the terminal discrimination unit;

[0008] After obtaining the fall judgment information, the terminal discrimination unit performs credibility analysis on the fall judgment information, obtains a fall judgment result through the credibility analysis, and sends a call control instruction to the two-way communication unit according to the fall judgment result;

[0009] After receiving the call control instruction, the two-way communication unit generates a call reminder to the management platform and feeds back a call connection instruction through the management platform to connect the two-way call;

[0010] The communication guarantee unit detects the voice input through the management platform, and when detecting the existence of voice input, records it as a call connection signal. The communication guarantee unit obtains the voice output in the home environment through the identification information acquisition unit, compares the voice output with the call connection signal, and generates a call normal signal or a call abnormal signal according to the comparison result;

[0011] The alarm trigger unit obtains the fall judgment result through the terminal discrimination unit, automatically generates a warning signal and sends it to the management platform. The alarm trigger unit obtains the call abnormal signal through the communication guarantee unit and conducts a call abnormal warning through the management platform.

[0012] As a preferred embodiment of the present invention, the fall judgment information collected by the identification information acquisition unit includes millimeter wave radar data, image data, and sound data;

[0013] After obtaining the millimeter wave radar data, the identification information acquisition unit performs point cloud imaging and target tracking processing on the millimeter wave radar data, sets a plurality of control points within the target range after target tracking, conducts continuous coordinate analysis on each control point, thereby obtaining the motion trajectory of each control point, and thus obtaining the motion speed and drop distance of different control points.

[0014] As a preferred embodiment of the present invention, when analyzing the image data, the identification information acquisition unit performs personnel identification on the collected image data, continuously tracks the identified personnel, and intelligently identifies the current action posture of the personnel through an algorithm to obtain the personnel dynamic probability, and sends the personnel dynamic probability to the terminal discrimination unit;

[0015] After the identification information collection unit collects the sound data, it compares the collected sound data with the set sound model, obtains the overlap between the sound data and the fall sound model, the call sound model and other sound models based on the comparison result, and sends the overlap to the terminal identification unit.

[0016] As a preferred embodiment of the present invention, after obtaining the drop distance of the control point, the terminal identification unit compares the drop distance with the set height drop threshold. If the drop distance is greater than the set height drop threshold, the movement speed is compared. If the movement speed is greater than the set movement speed, it is determined to be a fall control point. If the drop distance is not greater than the set height drop threshold or the movement speed is not greater than the set movement speed, it is determined to be a normal control point.

[0017] After the terminal identification unit has classified all the control points, the proportion of the falling control points in all the control points is calculated to obtain the falling probability proportion.

[0018] As a preferred embodiment of the present invention, the dynamic probability of the person obtained by the terminal identification unit includes a normal probability and a fall probability, and the terminal identification unit obtains the coincidence degree of the sound data and simultaneously obtains the fall probability ratio of the fall control point;

[0019] The terminal identification unit records the fall probability, the overlap of the fall sound model and the fall probability ratio as Qd, Qs and Qk respectively, and assigns different weight coefficients α, β and θ respectively, and performs weighted averaging through the formula to obtain the final fall credibility;

[0020] The terminal judgment unit compares the fall credibility with the threshold. If the fall credibility is greater than or equal to the set threshold, a fall signal is generated. If the fall credibility is less than the set threshold, a suspected fall signal is generated. After the fall signal is generated, a call control instruction is immediately generated.

[0021] As a preferred embodiment of the present invention, the two-way communication unit controls the recording and broadcasting components in the home environment, and is connected to the management platform through a wired network and a mobile network to achieve two-way communication;

[0022] After generating a call reminder to the management platform, the two-way call unit starts timing and compares the timing time with the set gear, and makes reminders of different degrees according to the gear to which the timing belongs. If a call connection instruction is fed back by the management platform, the timing is ended.

[0023] As a preferred embodiment of the present invention, after the communication assurance unit obtains the sound input through the management platform, it recognizes the content of the sound input through the sound recognition algorithm and records the recognized content. At the same time, after the communication assurance unit obtains the sound output in the home environment through the recognition information collection unit, it also recognizes the output content of the sound, and compares the content recognized by the sound input and the sound output to obtain the conversion coincidence degree;

[0024] The communication guarantee unit directly compares the waveforms of the sound input and the sound output, and obtains the waveform coincidence degree according to the waveform comparison;

[0025] The communication guarantee unit performs a comprehensive evaluation based on the waveform overlap and the conversion overlap, determines whether the waveform overlap and the conversion overlap meet the standard requirements, and obtains a normal sound output signal or an abnormal sound output signal.

[0026] As a preferred embodiment of the present invention, after the alarm triggering unit obtains the suspected fall signal through the terminal identification unit, it sends the video image to the management platform, reminds through the management platform, and manually feeds back a confirmation signal through the management platform.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. In the present invention, possible falls in the home environment are collected simultaneously by millimeter wave radar, sound detection and image acquisition, and a comprehensive judgment is made based on the collected information, thereby improving the accuracy of the judgment on the fall situation and avoiding missed or misjudged falls, thereby improving the response speed of detection of falls in the elderly living alone and ensuring the safety of the detected elderly.

[0029] 2. In the present invention, after a fall occurs, the system automatically accesses the two-way call function, thereby realizing two-way call across the air with the help of fixed sound collection components and sound playback components. When an elderly person falls, the guardian or management personnel can be contacted in time, so that the fallen elderly person can be rescued in time to avoid danger caused by falling.

[0030] 3. In the present invention, the two-way call system can verify the two-way call, so as to ensure that the two-way call system is stable and effective during operation. When the operation of the two-way call system is abnormal, it can detect and feedback the signal in time, so that the guardian or manager can take other remedial measures to ensure the safety of the elderly after an accidental fall. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0032] Figure 1 is the system block diagram of the present invention;

[0033] Figure 2 is the system flow chart of the present invention. Specific Embodiments

[0034] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts belong to the scope of protection of the present invention.

[0035] Embodiment 1:

[0036] Please refer to Figure 1 - Figure 2 As shown, a home fall detection system based on two-way communication over the Internet of Things includes an identification information acquisition unit, a terminal discrimination unit, a two-way communication unit, a communication guarantee unit, and an alarm trigger unit. The identification information acquisition unit is used to collect fall judgment information and send the collected fall judgment information to the terminal discrimination unit. The fall judgment information collected by the identification information acquisition unit includes millimeter-wave radar data, image data, and sound data;

[0037] After obtaining the millimeter-wave radar data, the identification information acquisition unit performs point cloud imaging and target tracking processing on the millimeter-wave radar data, and sets multiple control points within the target range after target tracking, and continuously analyzes the coordinates of each control point to obtain the motion trajectory of each control point, thereby obtaining the motion speed and drop distance of different control points;

[0038] When analyzing the image data, the identification information acquisition unit performs human identification on the collected image data, continuously tracks the identified person, and intelligently identifies the current action posture of the person through an algorithm to obtain the human dynamic probability, and sends the human dynamic probability to the terminal discrimination unit;

[0039] After collecting the sound data, the identification information acquisition unit compares the collected sound data with the set sound model, obtains the coincidence degree of the sound data with the fall sound model, the call sound model, and other sound models according to the comparison result, and sends the coincidence degree to the terminal discrimination unit;

[0040] After obtaining the fall judgment information, the terminal discrimination unit performs credibility analysis on the fall judgment information, obtains the fall judgment result through the credibility analysis, and sends a call control instruction to the two-way communication unit according to the fall judgment result;

[0041] After obtaining the elevation difference distance of the control point, the terminal discrimination unit compares the elevation difference distance with the set height elevation threshold. If the elevation difference distance is greater than the set height elevation threshold, it compares the movement speed. If the movement speed is greater than the set movement speed, it is determined as a fall control point. If the elevation difference distance is not greater than the set height elevation threshold or the movement speed is not greater than the set movement speed, it is determined as a normal control point. The height elevation threshold is determined by the height of the person detected by the fall detection system, and 0.4 to 0.7 times the height of the detected person is selected.

[0042] After the terminal discrimination unit classifies all the control points, it calculates the proportion of the fall control points among all the control points to obtain the proportion of the fall probability.

[0043] The personnel dynamic probability obtained by the terminal discrimination unit includes the normal probability and the fall probability. The terminal discrimination unit obtains the coincidence degree of the sound data and simultaneously obtains the proportion of the fall probability of the fall control points.

[0044] The terminal discrimination unit records the fall probability, the coincidence degree of the fall sound model, and the proportion of the fall probability as Qd, Qs, and Qk respectively, and simultaneously assigns different weight coefficients α, β, and θ, and performs weighted averaging through the formula to obtain the final fall credibility KD. ;

[0045] The terminal judgment unit compares the fall credibility with the threshold. If the fall credibility is greater than or equal to the set threshold, it generates a fall signal. If the fall credibility is less than the set threshold, it generates a suspected fall signal. After generating the fall signal, it immediately generates a call control instruction. After receiving the call control instruction, the two-way call unit generates a call reminder to the management platform and connects the two-way call through the call connection instruction feedback by the management platform. The set threshold is the empirical data obtained from the experiment in the experimental environment, and a certain proportion of the data is selected through statistical methods.

[0046] The two-way call unit controls the recording and broadcasting components in the home environment and is connected to the management platform through the wired network and the mobile network. The management platform connects to the management personnel through the network to realize two-way calls. The way for the management personnel to make two-way calls can be the mobile phone APP or the PC-side central control software.

[0047] Embodiment 2:

[0048] Please refer to Figure 1 - Figure 2 as shown in

[0049] After generating a call reminder to the management platform, the two-way call unit starts timing, compares the timing time with the set gears, and gives different levels of reminders according to the gear to which the timing belongs. If a call connection instruction is received from the management platform, the timing ends;

[0050] The communication guarantee unit detects the voice input through the management platform, and records it as a call connection signal when voice input is detected. The communication guarantee unit obtains the voice output in the home environment through the identification information collection unit. If there is voice output in the home environment, a normal call signal is generated. If there is no voice output in the home environment, an abnormal call signal is generated. The communication guarantee unit sends the abnormal call signal and the normal call signal to the management platform through the network;

[0051] After obtaining the voice input through the management platform, the communication guarantee unit identifies the content of the voice input through a voice recognition algorithm and records the recognized content. At the same time, after obtaining the voice output in the home environment through the identification information collection unit, the communication guarantee unit also identifies the output content of the voice, and compares the content recognized from the voice input and the voice output to obtain the conversion coincidence degree;

[0052] The communication guarantee unit directly compares the waveforms of the voice input and the voice output, and obtains the waveform coincidence degree according to the waveform comparison;

[0053] The communication guarantee unit makes a comprehensive evaluation based on the waveform coincidence degree and the conversion coincidence degree, judges whether the waveform coincidence degree and the conversion coincidence degree meet the standard requirements, and obtains a normal voice output signal or an abnormal voice output signal;

[0054] After obtaining a fall signal through the terminal discrimination unit, the alarm trigger unit automatically generates a warning signal and sends it to the management platform. After obtaining a suspected fall signal through the terminal discrimination unit, the alarm trigger unit sends the video image to the management platform and gives a reminder through the management platform. By manually feeding back a confirmation signal through the management platform, the alarm trigger unit obtains an abnormal call signal through the communication guarantee unit and gives an abnormal call warning through the management platform.

[0055] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the present invention, so that those skilled in the art in the relevant technical field can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A home fall detection system based on two-way communication of the Internet of Things, characterized in that: It includes an identification information collection unit, a terminal identification unit, a two-way conversation unit, a communication guarantee unit and an alarm triggering unit, wherein the identification information collection unit is used to collect fall judgment information and send the collected fall judgment information to the terminal identification unit; After acquiring the fall judgment information, the terminal identification unit performs a credibility analysis on the fall judgment information, obtains a fall judgment result through the credibility analysis, and sends a call control instruction to the two-way call unit according to the fall judgment result; After receiving the call control instruction, the two-way call unit generates a call reminder to the management platform, and feeds back a call connection instruction through the management platform to connect the two-way call; The communication guarantee unit detects the sound input through the management platform, and when the sound input is detected, records it as a call connection signal. The communication guarantee unit obtains the sound output in the home environment through the identification information collection unit, and compares the sound output with the call connection signal, and generates a normal call signal or an abnormal call signal according to the comparison result; The alarm triggering unit obtains the fall judgment result through the terminal identification unit, and automatically generates an early warning signal and sends it to the management platform. The alarm triggering unit obtains the call abnormality signal through the communication guarantee unit, and issues a call abnormality early warning through the management platform.

2. The home fall detection system based on two-way communication of the Internet of Things according to claim 1 is characterized in that: The fall judgment information collected by the identification information collection unit includes millimeter wave radar data, image data and sound data; After obtaining the millimeter-wave radar data, the identification information acquisition unit performs point cloud imaging and target tracking processing on the millimeter-wave radar data, and sets multiple control points within the target range after target tracking, and performs continuous coordinate analysis on each control point to obtain the motion trajectory of each control point, thereby obtaining the motion speed and drop distance of different control points.

3. The home fall detection system based on two-way communication of the Internet of Things according to claim 1 is characterized in that: When analyzing the image data, the identification information acquisition unit identifies the person in the collected image data, continuously tracks the identified person, and intelligently identifies the person's current action posture through an algorithm to obtain the person's dynamic probability, and sends the person's dynamic probability to the terminal identification unit; After the identification information collection unit collects the sound data, it compares the collected sound data with the set sound model, obtains the overlap between the sound data and the fall sound model, the call sound model and other sound models based on the comparison result, and sends the overlap to the terminal identification unit.

4. The home fall detection system based on two-way communication of the Internet of Things according to claim 2 is characterized in that: After obtaining the drop distance of the control point, the terminal identification unit compares the drop distance with the set height drop threshold. If the drop distance is greater than the set height drop threshold, the movement speed is compared. If the movement speed is greater than the set movement speed, it is determined to be a fall control point. If the drop distance is not greater than the set height drop threshold or the movement speed is not greater than the set movement speed, it is determined to be a normal control point. After the terminal identification unit has classified all the control points, the proportion of the falling control points in all the control points is calculated to obtain the falling probability proportion.

5. The home fall detection system based on two-way communication of the Internet of Things according to claim 3 is characterized in that: The dynamic probability of the person obtained by the terminal identification unit includes a normal probability and a fall probability. The terminal identification unit obtains the coincidence degree of the sound data and simultaneously obtains the fall probability ratio of the fall control point; The terminal identification unit records the fall probability, the overlap of the fall sound model and the fall probability ratio as Qd, Qs and Qk respectively, and assigns different weight coefficients α, β and θ respectively, and performs weighted averaging through the formula to obtain the final fall credibility; The terminal judgment unit compares the fall credibility with the threshold. If the fall credibility is greater than or equal to the set threshold, a fall signal is generated. If the fall credibility is less than the set threshold, a suspected fall signal is generated. After the fall signal is generated, a call control instruction is immediately generated.

6. The home fall detection system based on two-way communication of the Internet of Things according to claim 1 is characterized in that: The two-way communication unit controls the recording and broadcasting components in the home environment and connects with the management platform through a wired network and a mobile network to achieve two-way communication; After generating a call reminder to the management platform, the two-way call unit starts timing and compares the timing time with the set gear, and makes reminders of different degrees according to the gear to which the timing belongs. If a call connection instruction is fed back by the management platform, the timing is ended.

7. The home fall detection system based on two-way communication of the Internet of Things according to claim 1 is characterized in that: After the communication assurance unit obtains the sound input through the management platform, it recognizes the content of the sound input through the sound recognition algorithm and records the recognized content. At the same time, after the communication assurance unit obtains the sound output in the home environment through the recognition information collection unit, it also recognizes the output content of the sound, and compares the recognized content of the sound input and the sound output to obtain the conversion coincidence degree; The communication guarantee unit directly compares the waveforms of the sound input and the sound output, and obtains the waveform coincidence degree according to the waveform comparison; The communication guarantee unit performs a comprehensive evaluation based on the waveform overlap and the conversion overlap, determines whether the waveform overlap and the conversion overlap meet the standard requirements, and obtains a normal sound output signal or an abnormal sound output signal.

8. The home fall detection system based on two-way communication of the Internet of Things according to claim 1 is characterized in that: After the terminal identification unit obtains the suspected fall signal, the alarm trigger unit sends the video image to the management platform, and reminds through the management platform, and manually feeds back a confirmation signal through the management platform.

Images