A digital-based smart elderly care monitoring system

Abstract

The present invention discloses a digital-based intelligent elderly care monitoring system, which belongs to the technical field of elderly care monitoring, and comprises a body monitoring module, a scene monitoring module and a monitoring module; the body monitoring module is used for performing body monitoring on a monitoring target, obtaining a body assessment result corresponding to the monitoring target, and performing real-time feedback and sharing of the body assessment result; the monitoring module is used for performing data sharing with the monitoring personnel of the corresponding monitoring target; the scene monitoring module performs video monitoring on each monitoring target in a nursing home, and obtains monitoring videos of the monitoring targets; performs desensitization processing on the monitoring videos, and obtains target videos; shares the target videos with the monitoring module; generates a heart rate change curve according to the heart rate data of the monitoring target; performs real-time abnormality recognition according to the monitoring videos and the heart rate change curve, and obtains corresponding behavior assessment results and emotion assessment results; and shares the behavior assessment results and emotion assessment results with the corresponding monitoring modules.

Description

Technical Field

[0001] The present invention belongs to the technical field of elderly care monitoring, and specifically is a digital-based intelligent elderly care monitoring system. Background Art

[0002] As the trend of population aging becomes increasingly serious, nursing homes, as important elderly care service institutions, are facing increasing pressure and challenges. Traditional nursing home management methods often rely on manual daily care and health management of the elderly, which has many limitations, such as low efficiency and slow response speed. In particular, staff bullying of the elderly often occurs, causing many elderly family members to be afraid of sending the elderly to nursing homes for care. Based on this, in order to improve elderly care services and enhance the trust of the elderly family members, the present invention provides a digital-based intelligent elderly care monitoring system. Summary of the invention

[0003] In order to solve the problems existing in the above solutions, the present invention provides a digital-based smart elderly care monitoring system.

[0004] The purpose of the present invention can be achieved by the following technical solutions:

[0005] A digital-based smart elderly care monitoring system, including a body monitoring module, a scene monitoring module and a monitoring module;

[0006] The body monitoring module is used to perform body monitoring on the monitoring target, obtain disease data of each monitoring target, and set corresponding body monitoring standards for each monitoring standard according to the disease data;

[0007] Acquire the physical monitoring data of the monitoring target in real time; set the monitoring value corresponding to each monitoring item according to the physical monitoring data; set the corresponding monitoring standard value for each monitoring item according to the physical monitoring standard and the monitoring value; evaluate the monitoring target according to the monitoring standard value and the monitoring value, obtain the physical evaluation result corresponding to the monitoring target, and provide real-time feedback and share the physical evaluation result.

[0008] Furthermore, the method for evaluating the monitoring target according to the monitoring standard value and the monitoring value includes:

[0009] The monitoring value is marked as Gi, i represents the corresponding monitoring item, i = 1, 2, ..., n, n is a positive integer; the monitoring standard value is marked as GBi;

[0010] According to the formula Calculate the corresponding physical assessment value;

[0011] Where: GQ is the physical assessment value, βi is the preset proportional coefficient of the corresponding monitoring item, 0<βi≤1;

[0012] When the body assessment value is greater than the threshold value X1, the body monitoring is assessed to be abnormal;

[0013] When the body evaluation value is not greater than the threshold value X1, the body monitoring is evaluated to be normal.

[0014] The monitoring module is used to share data with the monitoring personnel of the corresponding monitoring targets.

[0015] The scene monitoring module performs video monitoring on each monitoring target in the nursing home and establishes a monitoring model, in which the monitoring range and location of each monitoring device are marked;

[0016] Acquire the monitoring target location in real time, and dynamically mark it in the monitoring model according to the identified monitoring target location; identify the monitoring device corresponding to the monitoring target location, and obtain the monitoring video of the corresponding monitoring device; desensitize the monitoring video to obtain the target video corresponding to the monitoring target; share the target video with the corresponding monitoring module;

[0017] Acquire the heart rate data of the monitored target in real time, and generate the corresponding heart rate change curve according to the acquired heart rate data;

[0018] Real-time abnormality recognition is performed based on the monitoring video and heart rate change curve to obtain corresponding behavior assessment results and emotion assessment results; the behavior assessment results and emotion assessment results are shared with the corresponding monitoring modules.

[0019] Furthermore, the method for real-time abnormality recognition based on monitoring video and heart rate change curve includes:

[0020] Establish an abnormal behavior recognition model. The expression of the abnormal behavior recognition model is: Where: s is the surveillance video; the output data is the behavior abnormality value 1 or 0;

[0021] Performing real-time analysis on the surveillance video through the abnormal behavior recognition model to obtain corresponding abnormal behavior values; generating corresponding abnormal behavior curves according to each of the abnormal behavior values;

[0022] In the behavior abnormality curve, a line segment with a behavior abnormality value of 1 is identified in real time and marked as a behavior abnormality segment; each behavior abnormality segment is combined and analyzed to obtain each behavior analysis segment;

[0023] Identify the time span corresponding to each behavior analysis segment and mark it as behavior duration;

[0024] Obtain a heart rate change curve, and match the heart rate change curve with a current abnormal behavior curve; divide the heart rate change curve into a heart rate verification curve and an emotion analysis curve according to the abnormal behavior curve;

[0025] Perform behavior verification on the corresponding behavior analysis segment according to the heart rate verification curve to obtain the corresponding behavior evaluation result;

[0026] Perform sentiment analysis according to the sentiment analysis curve to obtain the corresponding sentiment evaluation results.

[0027] Furthermore, the method for merging and analyzing each abnormal behavior segment includes:

[0028] Identify the interval time between each adjacent abnormal behavior segment, and compare the obtained interval time with the threshold value X2;

[0029] When the interval time is greater than the threshold X2, the corresponding two abnormal behavior segments are not merged;

[0030] When the interval time is not greater than the threshold value X2, the corresponding two abnormal behavior segments are merged to obtain a new abnormal behavior segment;

[0031] The same process is repeated until no adjacent abnormal behavior segments meet the merging requirements, and each abnormal behavior segment is marked as a behavior analysis segment.

[0032] Furthermore, the method of performing behavior verification on the corresponding behavior analysis segment according to the heart rate verification curve includes:

[0033] Real-time identification of the heart rate change value of the heart rate calibration curve, that is, the heart rate value that has changed compared with the previous time;

[0034] According to the formula WXPt = BLt × YCt × 1.1 t-t0 Calculate the behavior check value of the corresponding time; where: WXPt is the behavior check value, t is time; BLt is the heart rate change value; YCt is the behavior abnormality value, t0 is the initial time of the current behavior analysis segment;

[0035] When the behavior check value is greater than the threshold value X3, the behavior check is evaluated as abnormal;

[0036] When the behavior check value is not greater than the threshold value X3, the behavior check is evaluated to be normal.

[0037] Furthermore, the method for performing sentiment analysis according to the sentiment analysis curve includes:

[0038] Set corresponding emotion analysis data according to the emotion analysis curve and the surveillance video; analyze the emotion analysis data through the emotion anomaly assessment model to obtain the corresponding heart rate anomaly value;

[0039] Generate a corresponding abnormal heart rate curve according to the obtained abnormal heart rate value; fit the abnormal heart rate curve to obtain a corresponding abnormal heart rate function, and mark the abnormal heart rate function as XT(t), where t is time;

[0040] Set the unit time, and determine the time interval corresponding to the unit time in real time based on the current time, marked as [T1, T2], where T2 is the current time;

[0041] According to the formula Calculate the corresponding emotion value; where: PQE is the emotion value;

[0042] The corresponding emotion evaluation result is determined according to the emotion value.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] Through the mutual cooperation between the body monitoring module, scene monitoring module and monitoring module, intelligent monitoring of the elderly can be realized and the health status of the elderly can be understood in time; at the same time, dynamic monitoring of the living conditions of the elderly can be realized to ensure the physical and mental health of the elderly, and the corresponding evaluation data can be shared with the elderly's guardians, so that the guardians can truly understand the actual situation of the elderly. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 This is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] like Figure 1 As shown, a digital-based smart elderly care monitoring system includes a body monitoring module, a scene monitoring module and a monitoring module;

[0049] The body monitoring module is used to use various existing smart wearable devices (such as smart bracelets, smart watches, etc.) to monitor the elderly's body, obtain corresponding body monitoring data, such as heart rate, blood pressure, blood sugar, body temperature, etc., and mark the corresponding elderly as monitoring targets;

[0050] Obtain the disease data of each monitoring target, that is, whether the monitoring target currently has the corresponding disease, such as various chronic diseases; set the corresponding physical monitoring standards for the monitoring target according to the obtained disease data; the specific standards are set by the corresponding professionals according to the actual conditions of different elderly people under different diseases, forming different reference standards, and then establishing the corresponding disease case library, matching the disease case library according to the disease data of the monitoring target, and obtaining the corresponding physical monitoring standards;

[0051] According to the data corresponding to each monitoring item in the body monitoring data, the monitoring value corresponding to each monitoring item is set, such as the value corresponding to the monitoring items such as temperature and heart rate;

[0052] According to the monitoring value of each monitoring item, the corresponding monitoring standard value is matched from the physical monitoring standard. For example, if the standard interval of a monitoring item is [a1, a2], if the corresponding monitoring value is greater than a2, then the monitoring standard value is a2; if it is within the interval, then the monitoring standard value is the same as the monitoring value, that is, the monitoring standard value is the closest value to the monitoring value in the standard interval;

[0053] The monitoring value is marked as Gi, i represents the corresponding monitoring item, i = 1, 2, ..., n, n is a positive integer; the monitoring standard value is marked as GBi;

[0054] According to the formula Calculate the corresponding physical assessment value; where: GQ is the physical assessment value, βi is the preset proportional coefficient of the corresponding monitoring item, 0<βi≤1;

[0055] When the body assessment value is greater than the preset value, the body monitoring is assessed to be abnormal;

[0056] When the body assessment value is not greater than the preset value, the body monitoring is assessed to be normal.

[0057] Provide real-time feedback and share the physical assessment results; such as sharing with staff, elderly people’s family members, etc.

[0058] The monitoring module is used by the guardian of the corresponding monitoring target, who is generally the elderly person's family members or other guardians, and is used to display the corresponding physical assessment results, monitoring equipment and other related data to the guardian.

[0059] The scene monitoring module is used to perform video monitoring of each monitoring target in the nursing home, establish a corresponding three-dimensional visualization model according to the building information of the nursing home, and mark the corresponding three-dimensional visualization model according to the monitoring range and position of each monitoring device, so as to facilitate the subsequent rapid identification of the corresponding elderly according to the elderly positioning and monitoring video; avoid using video analysis for identification, improve data processing efficiency and recognition efficiency; mark the current three-dimensional visualization model as a monitoring model;

[0060] The location of the monitoring target can be obtained in real time, and the location can be determined based on the smart wearable device of the monitoring target; the corresponding mark and dynamic update are made in the monitoring model according to the identified monitoring target location; the monitoring device corresponding to the monitoring target location is identified, and the monitoring video of the corresponding monitoring device is obtained; the monitoring video is desensitized to obtain the target video corresponding to the monitoring target; the target video is shared with the corresponding monitoring module;

[0061] Acquire the heart rate data of the monitored target in real time, and generate the corresponding heart rate change curve according to the acquired heart rate data, with time as the horizontal axis and heart rate as the vertical axis;

[0062] Real-time anomaly recognition is performed based on surveillance video and heart rate change curve. The target video is used for sharing, but surveillance video is still used for anomaly analysis; corresponding behavior assessment results and emotion assessment results are obtained.

[0063] Among them, the surveillance video is desensitized, that is, other people and sounds that are not the monitored targets are removed to avoid infringing on the privacy of others, and the desensitization is performed using existing image processing technology.

[0064] Methods for real-time abnormality recognition based on surveillance video and heart rate change curves include:

[0065] Establish an abnormal behavior recognition model. The abnormal behavior recognition model is established based on the existing neural network. First, determine whether it is an abnormal behavior, and then output the corresponding abnormal behavior value according to the judgment result. Specifically, various abnormal behaviors are preset, such as nursing staff beating, fighting, quarreling, falling and other behaviors among the elderly, and the existing neural network technology is used for training and recognition. Or other image recognition technologies can be used for abnormal behavior recognition. The expression of the abnormal behavior recognition model is Where: s is the input data, i.e., the surveillance video; the output data is the behavior abnormality value 1 or 0; when it is determined that there is a behavior abnormality, it is also possible to determine what the corresponding abnormal behavior is as needed;

[0066] The monitoring video is analyzed in real time through the abnormal behavior recognition model to obtain the abnormal behavior value at the corresponding time; the corresponding abnormal behavior curve is generated according to the obtained abnormal behavior value; the horizontal axis is time and the vertical axis is the abnormal behavior value;

[0067] In the behavior abnormality curve, the line segment with the behavior abnormality value of 1 is identified in real time and marked as the behavior abnormality segment. An independent point is also regarded as the behavior abnormality segment; each behavior abnormality segment is combined and analyzed to obtain each behavior analysis segment;

[0068] Identify the time span corresponding to each behavior analysis segment and mark it as behavior duration;

[0069] Obtain the heart rate change curve, and correspond the heart rate change curve with the current abnormal behavior curve, that is, one-to-one correspondence according to time; divide the heart rate change curve into two parts according to the abnormal behavior curve, one is the heart rate change curve corresponding to each behavior analysis segment, marked as the heart rate verification curve; the other is the remaining heart rate change curve, marked as the emotion analysis curve;

[0070] The corresponding behavior analysis segment is behaviorally checked according to the heart rate calibration curve to obtain the corresponding behavior evaluation result. That is, if the check is not abnormal behavior, the behavior evaluation result is that the behavior check is normal; otherwise, the behavior check is abnormal and the corresponding abnormal behavior is marked.

[0071] Emotion analysis is performed according to the emotion analysis curve to obtain corresponding emotion evaluation results, including normal emotion evaluation and abnormal emotion evaluation.

[0072] The method of combining and analyzing each abnormal behavior segment includes:

[0073] Identify the interval time between each adjacent abnormal behavior segment, and compare the obtained interval time with the threshold value X2;

[0074] When the interval time is greater than the threshold X2, the corresponding two abnormal behavior segments are not merged;

[0075] When the interval time is not greater than the threshold value X2, the corresponding two abnormal behavior segments are merged to obtain a new abnormal behavior segment;

[0076] The same process is repeated until no adjacent abnormal behavior segments meet the merging requirements, and each abnormal behavior segment is marked as a behavior analysis segment.

[0077] The method of performing behavior verification on the corresponding behavior analysis segment according to the heart rate verification curve includes:

[0078] Real-time identification of the heart rate change value of the heart rate calibration curve, that is, the heart rate value that has changed compared with the previous time;

[0079] According to the formula WXPt = BLt × YCt × 1.1 t-t0 Calculate the behavior check value of the corresponding time; where: WXPt is the behavior check value, t is the time; BLt is the heart rate change value; YCt is the behavior abnormality value, t0 is the initial time of the current behavior analysis segment, that is, the time when it starts;

[0080] When the behavior check value is greater than the threshold value X3, the behavior check is evaluated as abnormal;

[0081] When the behavior check value is not greater than the threshold value X3, the behavior check is evaluated to be normal.

[0082] Methods for sentiment analysis based on sentiment analysis curves include:

[0083] Establish an abnormal emotion assessment model for analyzing the emotion analysis curve, mainly based on its heart rate change rate, heart rate value, and the language and facial image of the corresponding user. Specifically, it is established based on a CNN network or a DNN network, etc., and a corresponding training set is set for training. The training data includes input and output data. The input data includes the heart rate change rate, heart rate value, and the language and facial image of the corresponding user. The output data is the heart rate abnormality value. The heart rate abnormality value includes 0 and 1. When the assessment is abnormal, the heart rate abnormality value is 1, otherwise it is 0; train with the established training set, integrate the heart rate change rate, heart rate value, and the language and facial image of the corresponding user into emotion analysis data; analyze with the emotion abnormality assessment model after successful training to obtain the corresponding heart rate abnormality value;

[0084] Generate a corresponding abnormal heart rate curve according to the obtained abnormal heart rate value, with time as the horizontal axis and the abnormal heart rate value as the vertical axis; fit the abnormal heart rate curve to obtain a corresponding abnormal heart rate function, marked as XT(t), where t is time;

[0085] Set the unit time, that is, analyze based on one unit time, which is generally set to one hour, but can be adjusted accordingly as needed, such as 30 minutes, etc. The specific setting is done by professionals;

[0086] The time interval corresponding to the unit time is determined in real time based on the current time, marked as [T1, T2], where T2 is the current time;

[0087] According to the formula Calculate the corresponding emotion value; where: PQE is the emotion value;

[0088] When the emotion value is greater than the threshold value X4, the emotion evaluation is determined to be abnormal;

[0089] When the emotion value is not greater than the threshold value X4, the emotion evaluation is determined to be normal.

[0090] The above formulas are all calculated by removing dimensions and taking numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.

[0091] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A digital-based smart elderly care monitoring system, characterized in that: It includes body monitoring module, scene monitoring module and monitoring module; The body monitoring module is used to perform body monitoring on the monitoring targets, obtain disease data of each monitoring target, and set corresponding body monitoring standards for each monitoring standard according to the disease data; Acquire the physical monitoring data of the monitoring target in real time; set the monitoring value corresponding to each monitoring item according to the physical monitoring data; set the corresponding monitoring standard value for each monitoring item according to the physical monitoring standard and the monitoring value; evaluate the monitoring target according to the monitoring standard value and the monitoring value, obtain the physical evaluation result corresponding to the monitoring target, and provide real-time feedback and share the physical evaluation result; The monitoring module is used to share data with the monitoring personnel of the corresponding monitoring target; The scene monitoring module performs video monitoring on each monitoring target in the nursing home and establishes a monitoring model, in which the monitoring range and location of each monitoring device are marked; Acquire the monitoring target location in real time, and dynamically mark it in the monitoring model according to the identified monitoring target location; Identify the monitoring device corresponding to the monitoring target location and obtain the monitoring video of the corresponding monitoring device; Desensitize the surveillance video to obtain the target video corresponding to the monitored target; Share the target video to the corresponding monitoring module; Acquire the heart rate data of the monitored target in real time, and generate the corresponding heart rate change curve according to the acquired heart rate data; Perform real-time abnormal identification based on monitoring videos and heart rate change curves to obtain corresponding behavior assessment results and emotion assessment results; share the behavior assessment results and emotion assessment results with corresponding monitoring modules; Methods for real-time abnormality recognition based on surveillance video and heart rate change curves include: Establish an abnormal behavior recognition model, analyze the surveillance video in real time through the abnormal behavior recognition model, and obtain corresponding abnormal behavior values; generate corresponding abnormal behavior curves according to each of the abnormal behavior values; In the behavior abnormality curve, a line segment with a behavior abnormality value of 1 is identified in real time and marked as a behavior abnormality segment; each behavior abnormality segment is combined and analyzed to obtain each behavior analysis segment; Identify the time span corresponding to each behavior analysis segment and mark it as behavior duration; Obtain a heart rate change curve, and match the heart rate change curve with a current abnormal behavior curve; divide the heart rate change curve into a heart rate verification curve and an emotion analysis curve according to the abnormal behavior curve; Perform behavior verification on the corresponding behavior analysis segment according to the heart rate verification curve to obtain the corresponding behavior evaluation result; Perform sentiment analysis based on the sentiment analysis curve to obtain the corresponding sentiment evaluation results; The method of combining and analyzing each abnormal behavior segment includes: Identify the interval time between each adjacent abnormal behavior segment, and compare the obtained interval time with the threshold value X2; When the interval time is greater than the threshold X2, the corresponding two abnormal behavior segments are not merged; When the interval time is not greater than the threshold value X2, the corresponding two abnormal behavior segments are merged to obtain a new abnormal behavior segment; And so on, until no adjacent abnormal behavior segments meet the merging requirements, each abnormal behavior segment is marked as a behavior analysis segment; Methods for sentiment analysis based on sentiment analysis curves include: Set corresponding emotion analysis data according to the emotion analysis curve and the surveillance video; analyze the emotion analysis data through the emotion anomaly assessment model to obtain the corresponding heart rate anomaly value; Generate a corresponding abnormal heart rate curve according to the obtained abnormal heart rate value; fit the abnormal heart rate curve to obtain a corresponding abnormal heart rate function, and mark the abnormal heart rate function as XT(t), where t is time; Set the unit time, and determine the time interval corresponding to the unit time in real time based on the current time, marked as [T1, T2], where T1 and T2 are the two time endpoints before and after the corresponding unit time, and T2 is the current time; According to the formula Calculate the corresponding emotion value; where: PQE is the emotion value; The corresponding emotion evaluation result is determined according to the emotion value.

2. According to claim 1, a digital-based intelligent elderly care monitoring system is characterized in that: Methods for evaluating monitoring targets based on monitoring standard values ​​and monitoring values ​​include: The monitoring value is marked as Gi, i represents the corresponding monitoring item, i=1, 2, ..., n, n is a positive integer; the monitoring standard value is marked as GBi; According to the formula Calculate the corresponding physical assessment value; Where: GQ is the physical assessment value, βi is the preset proportional coefficient of the corresponding monitoring item, 0<βi≤1; When the body assessment value is greater than the threshold value X1, the body monitoring is assessed to be abnormal; When the body evaluation value is not greater than the threshold value X1, the body monitoring is evaluated to be normal.

3. According to claim 1, a digital-based intelligent elderly care monitoring system is characterized in that: The expression of the abnormal behavior recognition model is: ; Where: s is the surveillance video; the output data is the behavior abnormality value 1 or 0.

4. According to claim 1, a digital-based smart elderly care monitoring system is characterized in that: The method of performing behavior verification on the corresponding behavior analysis segment according to the heart rate verification curve includes: Real-time identification of the heart rate change value of the heart rate calibration curve, that is, the heart rate value that has changed compared with the previous time; According to the formula Calculate the behavior check value of the corresponding time; where: WXPt is the behavior check value, t is time; BLt is the heart rate change value; YCt is the behavior abnormality value, t0 is the initial time of the current behavior analysis segment; When the behavior check value is greater than the threshold value X3, the behavior check is evaluated as abnormal; When the behavior check value is not greater than the threshold value X3, the behavior check is evaluated to be normal.

Images