CN112183228B - Intelligent community endowment service system and method - Google Patents

Intelligent community endowment service system and method Download PDF

Info

Publication number
CN112183228B
CN112183228B CN202010937808.8A CN202010937808A CN112183228B CN 112183228 B CN112183228 B CN 112183228B CN 202010937808 A CN202010937808 A CN 202010937808A CN 112183228 B CN112183228 B CN 112183228B
Authority
CN
China
Prior art keywords
old
intelligent
old people
layer
community
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010937808.8A
Other languages
Chinese (zh)
Other versions
CN112183228A (en
Inventor
纪刚
周萌萌
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Qingdao Lianhe Chuangzhi Technology Co ltd
Original Assignee
Qingdao Lianhe Chuangzhi Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Qingdao Lianhe Chuangzhi Technology Co ltd filed Critical Qingdao Lianhe Chuangzhi Technology Co ltd
Priority to CN202010937808.8A priority Critical patent/CN112183228B/en
Publication of CN112183228A publication Critical patent/CN112183228A/en
Application granted granted Critical
Publication of CN112183228B publication Critical patent/CN112183228B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • G06Q50/265—Personal security, identity or safety
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/174—Facial expression recognition
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/51—Discovery or management thereof, e.g. service location protocol [SLP] or web services

Abstract

The invention discloses a community intelligent endowment service system and a community intelligent endowment service method.A community hardware acquisition system acquires human body posture data and thermal imaging images of an old person, matrix segmentation is carried out on the images, emotion probability distribution of the old person is obtained by applying emotion probability distribution network calculation, a BERT model is applied to forecast and evaluate the current physical and mental health of the old person, and a current safety risk index of the old person is fitted; and finally, obtaining the service requirements, suggestion guidance and article recommendation of the old people by adopting a clustering method. The old man observes the demand of oneself through the smart machine of wearing, acquires the service recommendation that intelligent platform gave through the old man intelligent terminal who carries, adopts a key formula operation, gets into the service that sharing platform obtained wanting. The invention realizes a service system of multiple intelligent services and safety guarantee, can solve the actual demand for the old, and is beneficial to solving the social problem of difficult nursing for the old.

Description

Intelligent community endowment service system and method
Technical Field
The invention relates to the technical field of big data, in particular to a community intelligent endowment service system and a community intelligent endowment service method.
Background
With the increasing aging of the population, the problem of aging becomes a social concern. At present, the home-based endowment mode accounts for 90 percent of the total number of people, the community endowment mode accounts for 7 percent, and the institution endowment mode only accounts for 3 percent. With the further increase of the age of the old people, specialized communities are necessary to provide home detection and service, and the community endowment mode promotes and supplements home endowment.
Traditional community endowment does not integrate informatization management, only depends on personnel to visit and investigate regularly, knows the life demand of the old people in the community, especially the widow old people, and provides help. The construction of 'internet plus' community endowment is a double-edged sword, on one hand, the problems of delay and block of traditional community endowment information are solved, on the other hand, the construction of the community internet needs high participation of the old, the operation flow on the internet is mastered, and for the old, the construction is a new challenge. For community group chat, although the community dynamics and question and answer can be conveniently mastered, the old people are required to check a large number of chat records, and when the number of people is large, one-to-one answer cannot be established, so that the old people are not beneficial to mastering the information which the old people want to know in real time.
Disclosure of Invention
In order to solve the technical problems, the invention provides a community intelligent endowment service system and a community intelligent endowment service method, so that the purposes of high-efficiency function aggregation, statistical analysis of physical and mental health of the old people, indication of safety risks, intelligent demand recommendation and realization of a service system with multiple intelligent services and safety guarantee through the analysis and prediction of emotion and behavior tracks of the old people are achieved.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a community intelligent endowment service system, comprising:
(1) the hardware acquisition system of the community: the intelligent monitoring equipment is provided with a thermal imaging and color cameras and is used for collecting thermal imaging and color image data of a specified old man, obtaining emotion probability distribution and behavior identification results of the old man through edge calculation, and uploading the results to a cloud service platform through edge-cloud communication; the intelligent equipment with positioning and sign monitoring functions worn by the old people uploads positioning data and sign data of the old people to the cloud service platform through edge-cloud communication;
(2) the cloud service platform calculates the obtained data to obtain physical and mental health conditions and safety risk indexes of the old, and then provides service requirements, suggestion guidance and article recommendation through the constructed intelligent platform:
(3) the intelligent platform is used for providing exclusive service requirements for specified old people, collecting and displaying data and information of the old people, structuring the data of the old people, and providing corresponding safety and service for different old people after effective information is obtained;
(4) old man intelligent terminal recommends for the service that every old man provided according to intelligent platform, adopts a key formula operation, need not the multistage service that can acquire wanting that seeks, and old man's key gets into can share platform information and service.
A community intelligent endowment service method comprises the following steps:
registering information of the old people participating in community intelligent care through an intelligent platform, synchronizing the information of the old people to a cloud service platform and a community hardware acquisition system, and acquiring a thermal imaging segmentation matrix of the appointed old people through the community hardware acquisition system;
designing an emotion probability distribution network based on edge calculation of intelligent monitoring equipment in a community hardware acquisition system, fusing the obtained thermal imaging segmentation matrix data of the old people with facial expression feature vectors obtained by the intelligent monitoring equipment as network input, calculating and obtaining emotion probability distribution of the old people, and synchronously uploading emotion probability distribution information of the old people to a cloud service platform through edge-cloud communication;
synchronizing the activity attributes of the old people obtained by the intelligent equipment in the community hardware acquisition system to a cloud service platform, combining the activity attributes of the old people with the emotion probability distribution information of the old people uploaded in the step two, wherein the attributes of the old people are respectively used as words to form sentences, inputting the sentences into a BERT model, predicting and evaluating the current physical and mental health of the old people, and fitting the current safety risk index of the old people;
step four, according to the current safety risk index evaluation result of the old, on a cloud service platform, a clustering method is adopted to obtain the service requirement, suggestion guidance and article recommendation of the old, and a cloud-edge communication mode is adopted to synchronize the data result to an intelligent platform and intelligent equipment;
and fifthly, the old people observe own requirements through the intelligent equipment worn by the old people, obtain service recommendation given by the intelligent platform through the intelligent terminal of the old people carried by the old people, and enter the sharing platform to obtain the desired service by one-key operation.
In the above scheme, the specific method of the first step is as follows:
(1) registering information of the old people participating in intelligent community care through an intelligent platform, and synchronizing the information of the old people to intelligent monitoring equipment and intelligent equipment in a cloud service platform and a community hardware acquisition system;
(2) acquiring human body posture data and thermal imaging images of the old according to intelligent monitoring equipment;
(3) traversing the thermal imaging image, and performing thermal value segmentation based on human body parts on the thermal imaging image according to human body posture data of the old people to obtain a thermal imaging segmentation matrix.
In the above scheme, the emotion probability distribution network in the second step is an LSTM time sequence network, and includes 5 layers of convolutional networks, where the convolutional kernel size of the first convolutional layer is 5 × 5 × 3, the number of feature layers is 16, the convolutional kernels of the other convolutional layers are 3 × 3 × 3, the number of feature layers is 32, 64, 128, 256, and 2 to 4 layers of convolutional layers are respectively followed by a bn layer, two full-connected layers are finally designed, the first full-connected layer outputs 512 emotion feature vectors, the second full-connected layer outputs 14 emotion classifications, and the softmax layer of the last layer is a predicted emotion probability distribution.
In the above scheme, in the third step, the BERT model uses a 12-layer transform structure for encoding, and the output of each layer of transform is the input of the next layer of transform, so as to encode a sentence.
In a further technical scheme, in the third step, a fusion layer + output layer network is designed, the doctor seeing situation, the interest situation and the abnormal situation of the old with the sample data concentrated and the output quantity coded by the BERT model are fused into the fusion layer network, 4 layers of transformers are adopted for fusion, the physical and mental health of the old is divided into excellent, good, common, bad and dangerous states, the probability intervals are 90-100, 70-90, 60-70, 40-60 and 0-40, the safety risk index is divided into safe, mild, moderate and severe states, the probability intervals are 80-100, 60-80, 30-60 and 0-30, 20 classifications are obtained after combination, and the classification of each old is finally obtained through calculation of the output layer.
Through the technical scheme, the community intelligent endowment service system and the community intelligent endowment service method have the following beneficial effects:
the method comprises the steps of collecting human body posture data and thermal imaging images of the old through a community hardware collection system, carrying out matrix segmentation on the images, obtaining emotion probability distribution of the old through emotion probability distribution network calculation, synchronously uploading emotion probability distribution information of the old to a cloud service platform through edge-cloud communication, predicting and evaluating the current physical and mental health of the old through a BERT model, and fitting a current safety risk index of the old; and finally, obtaining the service requirements, suggestion guidance and article recommendation of the old people by adopting a clustering method. The old man observes the demand of oneself through the smart machine of wearing, acquires the service recommendation that intelligent platform gave through the old man intelligent terminal who carries, adopts a key formula operation, gets into the service that sharing platform obtained wanting.
The invention realizes a service system of multiple intelligent services and safety guarantee, can solve the actual requirements for the old, provides different required services according to different personal conditions of the old, can quickly solve the problems of medical triage, life management and the emotion care of the old, does not omit or neglect, greatly reduces the pressure of community service personnel, enables the old to be in track with an information-based society without depending on family, improves the social sense and the value sense of the old through the interest education and the re-employment of the old, and is beneficial to solving the social problem of old care difficulty.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below.
FIG. 1 is a schematic diagram illustrating a community intelligent endowment service system according to an embodiment of the present invention;
FIG. 2 shows the emotion classifications 0-13 according to the present invention.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
The invention provides a community intelligent endowment service system, as shown in figure 1, comprising:
(1) the hardware acquisition system of the community:
the intelligent monitoring equipment with the thermal imaging and the color camera is used for collecting thermal imaging and color image data of the appointed old people, obtaining emotion probability distribution and behavior identification results of the old people through edge calculation, and uploading the results to the cloud service platform through edge-cloud communication.
The intelligent device with positioning and sign monitoring functions worn by the old people uploads positioning data and sign data of the old people to a cloud service platform through edge-cloud communication:
(a) collecting physical sign information 'heart rate, blood pressure, blood oxygen, body temperature and sleep' of a designated old man:
the method comprises two modes: (i) automatic measurement every 30 minutes, (ii) manual random measurement by the elderly; abnormal physical sign threshold values can be set in app or a platform, for example, the heart rate is not less than the low threshold value and not more than the high threshold value, the high threshold value and the low threshold value can be set according to specific conditions of the old, and other physical sign settings are similar. After the setting is finished, when the physical sign measurement value is not in the measurement interval, performing body sub-state reminding;
(b) the positioning information of appointed old people is collected, a UWB positioning mode is adopted indoors, GPS/Beidou/WiFi/LBS quadruple positioning is adopted outdoors, at the moment, the old people can be prevented from being out of range to cause danger by arranging a safety/danger fence, the old people have historical tracks, and the activity tracks and the step number statistics of the old people are recorded.
(2) The cloud service platform calculates the obtained data to obtain physical and mental health conditions and safety risk indexes of the old, and then provides service requirements, suggestion guidance and article recommendation through the constructed intelligent platform:
(3) the intelligent platform provides exclusive service demand for appointed old man, collects and demonstrates old man data and information, with old man data structurization, provides corresponding safety and service for different old man after obtaining effective information, specifically includes:
(i) inputting and displaying attribute information, physical sign monitoring information and safety monitoring information of the old people;
(ii) according to the physical and mental health open medical consultation and treatment and psychological consultation and treatment service of the old, the community staff are intelligently reminded to master the condition of the old and assist in treatment;
(iii) according to the physical condition and the safety risk index of the old people, the old people are intelligently reminded to reasonably arrange daily life and pay attention to safety by open ordering distribution and household service, old people college and entertainment activities, shopping and health care subscription and abnormal behavior alarming and processing service, and the old people with high safety risk index are subjected to high-frequency positioning and sign monitoring, so that the old people can freely arrange life and activities in a high-security community.
(4) The old man intelligent terminal recommends for the service that every old man provided according to intelligent platform, adopts a key formula operation, need not multistage seeking can obtain the service that wants, community bulletin, study content, recreational activity etc. the old man gets into one key and can share platform information and service.
A community intelligent endowment service method comprises the following steps:
step one, registering the information of the old people participating in community intelligent endowment through an intelligent platform, synchronizing the information of the old people to a cloud service platform and a community hardware acquisition system, and acquiring a thermal imaging segmentation matrix of the appointed old people through the community hardware acquisition system, wherein the thermal imaging segmentation matrix is as follows:
(1) through an intelligent platform, information of the Old people participating in the intelligent nursing of the community is registered, and the information set of the Old people is set to be { Old1,...,Oldi,...,OldMM is the number of the elderly in the set, Oldi＝{genderi,ageiThe ith old man, the gen deriAge being the sex of the elderlyiThe age of the old is 1-M, and the i is more than or equal to 1 and less than or equal to M, and the information of the old is synchronized to intelligent monitoring equipment and intelligent equipment in a cloud service platform and a community hardware acquisition system;
(2) collecting a human body posture data set and a thermal imaging image of the old according to the intelligent monitoring equipment, wherein the human body posture data set is expressed as follows
Figure BDA0002672557570000051
Means that the ith old man is monitored from t under the ith videoN+1 to tN+ a set of poses in time point, l denotes communityVideo monitoring number, tNRepresenting time nodes, 1-a representing time nodes tNSubsequent sequential discrete time points: t is tN+1≤tN+L≤tN+a，
Figure BDA0002672557570000052
Denotes the t-thNThe human body posture at the + L time point comprises a head, a chest, arms and legs, wherein,
Figure BDA0002672557570000053
representing the coordinates of the head posture joint points of the old man; the other parts have the same principle, wherein ind is more than or equal to 0 and less than or equal to 4; the thermographic image is represented as follows:
Figure BDA0002672557570000054
means that the ith old man is monitored from t under the ith videoN+1 to tNA thermographic image at + a time point;
(3) according to
Figure BDA0002672557570000055
In the area coordinate of (1), go through the t-thN+ L thermal imaging picture of old person that intelligent monitoring equipment gathered under time point
Figure BDA0002672557570000056
The corresponding matching area is used for solving the segmentation of the thermal imaging graph according to the following formula
Figure BDA0002672557570000057
Figure BDA0002672557570000058
Figure BDA0002672557570000059
Figure BDA00026725575700000510
Wherein the content of the first and second substances,
Figure BDA00026725575700000511
indicates t below the ith cameraN+LThermal imaging chart corresponding to the aged for the ith old at any moment
Figure BDA00026725575700000512
The segmentation comprises tN+LThe heat value of head, chest, arms and legs at the moment are collected by
Figure BDA00026725575700000513
It is shown that, among others,
Figure BDA0002672557570000061
the initial value is null, and when satisfied, the thermal imaging graph is traversed
Figure BDA0002672557570000062
Figure BDA0002672557570000063
When the temperature of the water is higher than the set temperature,
Figure BDA0002672557570000064
adding corresponding heat value in the thermal imaging image, (x, y) is corresponding thermal imaging image position coordinate, val _ cur (x, y) is the heat value under the position coordinate, and prop takes { Head, Chest, Arm, Leg } to respectively represent Head, Chest, Arm, Leg.
Designing an emotion probability distribution network based on edge calculation of intelligent monitoring equipment in a community hardware acquisition system, integrating the obtained thermal imaging segmentation matrix data of the old with facial expression feature vectors obtained by the intelligent monitoring equipment as network input, calculating and obtaining emotion probability distribution of the old, and synchronously uploading emotion probability distribution information of the old to a cloud service platform through edge-cloud communication, wherein the emotion probability distribution network specifically comprises the following steps:
(i) according to the intelligenceThe monitoring device can acquire the facial expression characteristics,
Figure BDA0002672557570000065
means that the ith old man is monitored from t under the ith videoN+1 to tN+ a facial expression feature set in time point, l represents community video monitoring number, tN+1≤tN+L≤tN+a，
Figure BDA0002672557570000066
Denotes the t-thN+ L expression feature vectors of the elderly at the time point;
(ii) probability distribution of emotion seeking:
designing an emotion probability distribution network, wherein the input quantity comprises tN+1～tNThe thermal imaging segmentation matrix and the facial expression feature vector sequence under + a belong to an LSTM time sequence network and comprise 5 layers of convolutional networks, wherein the size of a convolution kernel of a first layer of convolutional layer is 5 multiplied by 3, the number of feature layers is 16, the size of convolution kernels of other convolutional layers is 3 multiplied by 3, the number of feature layers is 32, 64, 128, 256 and 2-4 layers of convolutional layers are respectively provided with a bn layer at the back, two full-connection layers are finally designed, the first full-connection layer outputs 512 emotion feature vectors, the second full-connection layer outputs 14 emotion classifications, and the softmax layer of the last layer is predicted emotion probability distribution.
The input quantities are expressed as:
Figure BDA0002672557570000067
Figure BDA0002672557570000068
the emotion probability distribution network is represented as:
Figure BDA0002672557570000071
Figure BDA0002672557570000072
wherein cntl is the number of video monitoring, l is more than or equal to 1 and less than or equal to cntl represents the number of video monitoring, FaceiIndicates the ith old man is at tN+1≤tN+L≤tN+ a time face expression normalization matrix, Threml, under all video surveillanceiIndicates the ith old man is at tN+1≤tN+L≤tN+ a time thermal imaging segmentation normalization matrix under all video surveillance, li > 1 represents the number of network layers, alphalayer(l1)＝αconv(l1)，αlayer(li)＝[αconv(li)；αbn(li)]Is a network weight parameter vector, σlayer(li)＝[σconv(li),σbn(li)]Representing the network offset vector, outputting the 1 st convolutional layer as a set of features, outputting the 2-4 convolutional layers + BN layer as a set of features
Figure BDA0002672557570000073
The layer (li) represents the combined layer identifier, the formula (4) represents a relational formula of the output characteristics of the first layer and the input variables, and the formula (5) represents a calculation relationship between the characteristics of the two adjacent layers;
the emotional feature vector output by the first layer fully-connected layer is represented as:
Figure BDA0002672557570000074
Figure BDA0002672557570000075
wherein the content of the first and second substances,
Figure BDA0002672557570000076
representing the inde-th emotion characteristic value, wherein inde is more than or equal to 1 and less than or equal to 512, and obtaining the representation of emotion probability distribution according to the 5-th layer characteristic output:
Figure BDA0002672557570000077
Figure BDA0002672557570000078
wherein the content of the first and second substances,
Figure BDA0002672557570000079
the index is the characteristic value of the index emotional state, 0 is less than or equal to index and less than or equal to 13,
Figure BDA00026725575700000710
wherein the emotions corresponding to 0-13 sequentially include anger, fear, disgust, happiness, sadness, surprise, neutrality, anxiety, love, depression, slight, proud, perquisite and jealousy, and are specifically shown in fig. 2.
Step three, synchronizing the activity attributes of the old people obtained by the intelligent equipment in the community hardware acquisition system to a cloud service platform, combining the activity attributes and the emotion probability distribution information of the old people uploaded in the step two as words and sentences respectively, inputting the sentences into a BERT model, predicting and evaluating the current physical and mental health of the old people, and fitting the current safety risk index of the old people, wherein the details are as follows:
(1) setting an intelligent terminal to acquire an activity attribute data set of the old:
Figure BDA0002672557570000081
Figure BDA0002672557570000082
Figure BDA0002672557570000083
wherein lct0～lct4Respectively showing 5 positions (address, shopping, walking, sports, community activities),
Figure BDA0002672557570000084
respectively representing the positions of the old people, and taking values of 0 and 1, wherein when the old people appear at the positions, the value is 1, and otherwise, the value is 0;
Figure BDA0002672557570000085
respectively shows the times of the appearance of the old people in each position and the time length of each time,
Figure BDA0002672557570000086
other positions represent the same; range of each time of vital sign data
Figure BDA0002672557570000087
Wherein the content of the first and second substances,
Figure BDA0002672557570000088
indicating each occurrence at lct0The range of values of the heart rate at the location,
Figure BDA0002672557570000089
indicating each occurrence at lct0The value range of the blood pressure and the high pressure on the position,
Figure BDA00026725575700000810
indicating each occurrence at lct0The value range of the blood pressure and the low pressure on the position,
Figure BDA00026725575700000811
indicating each occurrence at lct0The value range of the blood oxygen at the position is
Figure BDA00026725575700000812
By
Figure BDA00026725575700000813
Next (heart rate range, blood pressure high pressure range, blood pressure low pressure range,blood oxygen range), other locations lct1～lct4The same;
Figure BDA00026725575700000814
wherein the content of the first and second substances,
Figure BDA00026725575700000815
indicating the first occurrence at lct0Under the position, the values of 9 behaviors (standing, walking, running, bending, squatting, sitting, falling, putting up, abnormal gathering) are 1 when each behavior occurs and 0 when no behavior occurs,
Figure BDA00026725575700000816
are shared by
Figure BDA00026725575700000817
Second at lct0Behavior value component under position, other position lct1～lct4The same is true.
Will old people activity attribute data WiEmotional probability EclsiAnd Old person attribute OldiAs a word, combine into a sentence: stci＝{Oldi,Wi,EclsiAnd f, collecting sample data of all the old people as follows: (Stcs) { Stc ═1,...,Stci,...,StccntpAnd i is more than or equal to 1 and less than or equal to cntp is the number of the old, and cntp is the number of the old in the sample set.
(2) Coding by using Stcs as the input of the BERT model and adopting a 12-layer transform structure in the BERT model, wherein the output of each layer of transform is the input of the next layer of transform, and is defined as
Figure BDA0002672557570000091
cnth represents the number of output elements per layer, 1 ≦ ii ≦ 12 represents the number of layers of the transformer, and Stcs is encoded;
(3) designing a fusion layer + output layer network, fusing the diagnosis condition, the interest condition and the abnormal condition of the old with sample data concentrated and the output quantity coded by a BERT network into a fusion layer network, adopting 4-layer transformers for fusion, dividing the physical and mental health of the old into { excellent, good, general, poor and dangerous }, combining probability intervals of { 90-100, 70-90, 60-70, 40-60 and 0-40 }, dividing the safety risk index into { safe, mild, moderate and severe }, and obtaining 20 classifications after the probability intervals are { 80-100, 60-80, 30-60 and 0-30 }, calculating through the output layer, and finally obtaining the classification of each old:
Figure BDA0002672557570000092
Figure BDA0002672557570000093
step four, according to the current safety risk index evaluation result of the old, on a cloud service platform, a clustering method is adopted to obtain the service requirement, suggestion guidance and article recommendation of the old, and a cloud-edge communication mode is adopted to synchronize the data result to an intelligent platform and an intelligent device, wherein the method specifically comprises the following steps:
(1) drawing up a service requirement array of the old:
serv _ demand ═ medical counseling, psychological counseling, takeaway, accompanying travel, safety monitoring,
shopping and distribution, health care, college of the old, employment of the old, entertainment activities,
housekeeping, care of entering the home, and care of living }
Assign a physical and mental health and safety risk index label for each service demand as shown in table 1:
TABLE 1 assigned wellness and safety Risk indices tags for each service requirement
Figure BDA0002672557570000094
Figure BDA0002672557570000101
(2) According to the evaluation result { p (health)i),p(safei) Clustering the service requirements to obtain a service requirement probability set of the ith old man:
Figure BDA0002672557570000102
Figure BDA0002672557570000103
representing the demand probability of the ith old man for the service demand type of CNTS, and representing the number of service demand types obtained by clustering, if
Figure BDA0002672557570000104
And then, the final service requirement of the old is obtained, and corresponding suggestion guidance and article recommendation in the database are obtained according to the service requirement, so that the intelligent community endowment service is realized.
And fifthly, the old people observe own requirements through the intelligent equipment worn by the old people, obtain service recommendation given by the intelligent platform through the intelligent terminal of the old people carried by the old people, and enter the sharing platform to obtain the desired service by one-key operation.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (4)

1. A community intelligent endowment service method adopts a community intelligent endowment service system, and is characterized in that the system comprises:
(1) the hardware acquisition system of the community: the intelligent monitoring equipment is provided with a thermal imaging and color cameras and is used for collecting thermal imaging and color image data of a specified old man, obtaining emotion probability distribution and behavior identification results of the old man through edge calculation, and uploading the results to a cloud service platform through edge-cloud communication; the intelligent equipment with positioning and sign monitoring functions worn by the old people uploads positioning data and sign data of the old people to the cloud service platform through edge-cloud communication;
(2) the cloud service platform calculates the obtained data to obtain physical and mental health conditions and safety risk indexes of the old, and then provides service requirements, suggestion guidance and article recommendation through the constructed intelligent platform:
(3) the intelligent platform is used for providing exclusive service requirements for specified old people, collecting and displaying data and information of the old people, structuring the data of the old people, and providing corresponding safety and service for different old people after effective information is obtained;
(4) the intelligent terminal for the old people adopts one-button operation according to the service recommendation provided by the intelligent platform for each old person, can acquire the desired service without multi-level searching, and can share the platform information and service when the old people enter the intelligent terminal by one button;
the method comprises the following steps:
registering information of the old people participating in community intelligent care through an intelligent platform, synchronizing the information of the old people to a cloud service platform and a community hardware acquisition system, and acquiring a thermal imaging segmentation matrix of the appointed old people through the community hardware acquisition system; the method comprises the following specific steps:
(1) through an intelligent platform, information of the Old people participating in the intelligent nursing of the community is registered, and the information set of the Old people is set to be { Old1,...,Oldi,...,OldMM is the number of the elderly in the set, Oldi＝{genderi,ageiThe ith old man, the gen deriAge being the sex of the elderlyiThe age of the old is 1-M, and the i is more than or equal to 1 and less than or equal to M, and the information of the old is synchronized to intelligent monitoring equipment and intelligent equipment in a cloud service platform and a community hardware acquisition system;
(2) collecting a human body posture data set and a thermal imaging image of the old according to the intelligent monitoring equipment, wherein the human body posture data set is expressed as follows
Figure FDA0003679251730000011
Means that the ith old man is monitored from t under the ith videoN+1 to tN+ a set of poses at time point, l denotes the number of video surveillance of the community, tNRepresents a time node, 1-a represents a time node tNSubsequent sequential discrete time points: t is tN+1≤tN+L≤tN+a，
Figure FDA0003679251730000012
Denotes the t-thNThe human body posture at the + L time point comprises a head, a chest, arms and legs, wherein,
Figure FDA0003679251730000013
representing the coordinates of the head posture joint points of the old man; the other parts have the same principle, wherein ind is more than or equal to 0 and less than or equal to 4; the thermographic image is represented as follows:
Figure FDA0003679251730000021
means that the ith old man is monitored from t under the ith videoN+1 to tNA thermographic image at + a time point;
(3) according to
Figure FDA0003679251730000022
In the area coordinate of (1), go through the t-thNOld people thermal imaging image acquired by intelligent monitoring equipment at + L time point
Figure FDA0003679251730000023
The corresponding matching area is used for solving the segmentation of the thermal imaging graph according to the following formula
Figure FDA0003679251730000024
Figure FDA0003679251730000025
Figure FDA0003679251730000026
Figure FDA0003679251730000027
Wherein the content of the first and second substances,
Figure FDA0003679251730000028
indicates t below the ith cameraN+LThermal imaging chart corresponding to the aged for the ith old at any moment
Figure FDA0003679251730000029
Including t, is used to generate a thermal imageN+LHeat value set of head, chest, arms and legs at time, using
Figure FDA00036792517300000210
It is shown that, among others,
Figure FDA00036792517300000211
the initial value is null, and as the thermal imaging map is traversed, when satisfied
Figure FDA00036792517300000212
Figure FDA00036792517300000213
When the temperature of the water is higher than the set temperature,
Figure FDA00036792517300000214
adding corresponding heat value in a thermal imaging image, (x, y) is corresponding thermal imaging image position coordinate, val _ cur (x, y) is the heat value under the position coordinate, and prop value is { Head, Chest, Arm, Leg } which respectively represents Head, Chest, Arm and Leg;
designing an emotion probability distribution network based on edge calculation of intelligent monitoring equipment in a community hardware acquisition system, fusing the obtained thermal imaging segmentation matrix data of the old people with facial expression feature vectors obtained by the intelligent monitoring equipment as network input, calculating and obtaining emotion probability distribution of the old people, and synchronously uploading emotion probability distribution information of the old people to a cloud service platform through edge-cloud communication;
the method comprises the following specific steps:
(i) according to the facial expression characteristics obtained by the intelligent monitoring equipment,
Figure FDA00036792517300000215
means that the ith old man is monitored from t under the ith videoN+1 to tN+ a facial expression feature set in time point, l represents community video monitoring number, tN+1≤tN+L≤tN+a，
Figure FDA00036792517300000216
Denotes the t-thN+ L expression feature vectors of the elderly at the time point;
(ii) probability distribution of emotion seeking:
designing an emotion probability distribution network, wherein the input quantity comprises tN+1～tNThe thermal imaging segmentation matrix and the facial expression feature vector sequence under + a belong to an LSTM time sequence network and comprise 5 layers of convolutional networks, wherein the size of a convolution kernel of a first layer of convolutional layer is 5 multiplied by 3, the number of feature layers is 16, the sizes of convolution kernels of other convolutional layers are 3 multiplied by 3, the number of feature layers is 32, 64, 128, 256 and 2-4 layers of convolutional layers are respectively provided with a bn layer at the back, finally two full-connection layers are designed, the first full-connection layer outputs 512 emotion feature vectors, the second full-connection layer outputs 14 emotion classifications, and the last softmax layer is predicted emotion probability distribution;
the input quantities are expressed as:
Figure FDA0003679251730000031
Figure FDA0003679251730000032
the emotion probability distribution network is represented as:
Figure FDA0003679251730000033
Figure FDA0003679251730000034
wherein cntl is the number of video monitoring, l is more than or equal to 1 and less than or equal to cntl represents the number of video monitoring, FaceiIndicates the ith old man is at tN+1≤tN+L≤tN+ a time face expression normalization matrix, Threml, under all video surveillanceiIndicates the ith old man is at tN+1≤tN+L≤tN+ a time thermal imaging segmentation normalization matrix under all video surveillance, li > 1 represents the number of network layers, alphalayer(l1)＝αconv(l1)，αlayer(li)＝[αconv(li)；αbn(li)]Is a network weight parameter vector, σlayer(li)＝[σconv(li),σbn(li)]Representing the network offset vector, outputting the 1 st convolutional layer as a set of features, outputting the 2-4 convolutional layers + BN layer as a set of features
Figure FDA0003679251730000035
The layer (li) represents the combined layer identifier, the formula (4) represents a relational formula of the output characteristics of the first layer and the input variables, and the formula (5) represents a calculation relationship between the characteristics of the two adjacent layers;
the emotional feature vector output by the first layer fully-connected layer is represented as:
Figure FDA0003679251730000036
Figure FDA0003679251730000037
wherein the content of the first and second substances,
Figure FDA0003679251730000041
expressing the inde-th emotion characteristic value, wherein inde is more than or equal to 1 and less than or equal to 512, and expressing emotion probability distribution according to the 5 th layer characteristic output:
Figure FDA0003679251730000042
Figure FDA0003679251730000043
wherein, the first and the second end of the pipe are connected with each other,
Figure FDA0003679251730000044
the index is the characteristic value of the index emotional state, 0 is less than or equal to index and less than or equal to 13,
Figure FDA0003679251730000045
wherein the emotions corresponding to 0-13 sequentially comprise anger, fear, disgust, happiness, sadness, surprise, neutrality, anxiety, love, depression, slight, proud, perquisite and jealousy;
synchronizing the activity attributes of the old people obtained by the intelligent equipment in the community hardware acquisition system to a cloud service platform, combining the activity attributes of the old people with the emotion probability distribution information of the old people uploaded in the step two, wherein the attributes of the old people are respectively used as words to form sentences, inputting the sentences into a BERT model, predicting and evaluating the current physical and mental health of the old people, and fitting the current safety risk index of the old people;
step four, according to the current safety risk index evaluation result of the old, on a cloud service platform, a clustering method is adopted to obtain the service requirement, suggestion guidance and article recommendation of the old, and a cloud-edge communication mode is adopted to synchronize the data result to an intelligent platform and intelligent equipment;
and fifthly, the old people observe own requirements through the intelligent equipment worn by the old people, obtain service recommendation given by the intelligent platform through the intelligent terminal of the old people carried by the old people, and enter the sharing platform to obtain the desired service by one-key operation.
2. The intelligent community endowment service method according to claim 1, wherein the specific method of the first step is as follows:
(1) registering information of the old people participating in intelligent community care through an intelligent platform, and synchronizing the information of the old people to intelligent monitoring equipment and intelligent equipment in a cloud service platform and a community hardware acquisition system;
(2) acquiring human body posture data and thermal imaging images of the old according to intelligent monitoring equipment;
(3) traversing the thermal imaging image, and performing thermal value segmentation based on human body parts on the thermal imaging image according to human body posture data of the old people to obtain a thermal imaging segmentation matrix.
3. The method as claimed in claim 1, wherein in the third step, the BERT model is encoded by using 12 layers of transform structures, and the output of each layer of transform is the input of the next layer of transform, so as to encode the sentence.
4. The intelligent community endowment service method according to claim 3, wherein in the third step, a fusion layer + output layer network is designed, sample data is concentrated on the doctor seeing, interest and abnormal situations of the old, and the output quantity coded by the BERT model is fused into the fusion layer network, 4 layers of transformers are adopted for fusion, the physical and mental health of the old is divided into excellent, good, common, poor and dangerous states, the probability intervals are 90-100, 70-90, 60-70, 40-60 and 0-40, the safety risk index is divided into safety, mild, moderate and severe states, the probability intervals are 80-100, 60-80, 30-60 and 0-30, 20 classifications are obtained after combination, and the classification of each old is obtained finally through calculation of the output layer.
CN202010937808.8A 2020-09-09 2020-09-09 Intelligent community endowment service system and method Active CN112183228B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010937808.8A CN112183228B (en) 2020-09-09 2020-09-09 Intelligent community endowment service system and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010937808.8A CN112183228B (en) 2020-09-09 2020-09-09 Intelligent community endowment service system and method

Publications (2)

Publication Number Publication Date
CN112183228A CN112183228A (en) 2021-01-05
CN112183228B true CN112183228B (en) 2022-07-08

Family

ID=73919982

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010937808.8A Active CN112183228B (en) 2020-09-09 2020-09-09 Intelligent community endowment service system and method

Country Status (1)

Country Link
CN (1) CN112183228B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112948019A (en) * 2021-03-11 2021-06-11 天津市智慧物联信息技术研究院 Endowment intelligent terminal service integration method based on edge computing architecture
CN112990787B (en) * 2021-05-10 2021-11-02 明品云(北京)数据科技有限公司 Service strategy determination method, system, equipment and medium
CN113270196B (en) * 2021-05-25 2023-07-14 郑州大学 Cerebral apoplexy recurrence risk perception and behavior decision model construction system and method
CN114840777B (en) * 2022-07-04 2022-09-27 杭州城市大脑有限公司 Multi-dimensional endowment service recommendation method and device and electronic equipment

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104184759A (en) * 2013-05-22 2014-12-03 天津市安维康家科技发展有限公司 Intelligent communality management service system
CN108717678A (en) * 2018-05-25 2018-10-30 青岛联合创智科技有限公司 A kind of wisdom endowment system
CN108900576A (en) * 2018-06-08 2018-11-27 上海驹源通信科技有限公司 A kind of intelligence community application service system based on 5G
CN110782379A (en) * 2019-11-04 2020-02-11 常州市小先信息技术有限公司 Smart home old-age care system scheme
CN111223498A (en) * 2020-01-10 2020-06-02 平安科技（深圳）有限公司 Intelligent emotion recognition method and device and computer readable storage medium
CN111374651A (en) * 2020-04-10 2020-07-07 江苏米孚致远物联科技有限公司 Novel wisdom endowment system

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105608642A (en) * 2015-12-15 2016-05-25 北京中科云集科技有限公司 Intelligent old age nursing industry service management method and system
CN107103568A (en) * 2017-04-28 2017-08-29 深圳市前海安测信息技术有限公司 Wisdom endowment information on services interactive system and method
CN111325946A (en) * 2020-01-19 2020-06-23 浙江理工大学 Fall detection method and system based on edge calculation
CN111240486B (en) * 2020-02-17 2021-07-02 河北冀联人力资源服务集团有限公司 Data processing method and system based on edge calculation
CN111590582A (en) * 2020-05-27 2020-08-28 华南理工大学 Intelligent medical service robot capable of realizing biofeedback and remote prospect

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104184759A (en) * 2013-05-22 2014-12-03 天津市安维康家科技发展有限公司 Intelligent communality management service system
CN108717678A (en) * 2018-05-25 2018-10-30 青岛联合创智科技有限公司 A kind of wisdom endowment system
CN108900576A (en) * 2018-06-08 2018-11-27 上海驹源通信科技有限公司 A kind of intelligence community application service system based on 5G
CN110782379A (en) * 2019-11-04 2020-02-11 常州市小先信息技术有限公司 Smart home old-age care system scheme
CN111223498A (en) * 2020-01-10 2020-06-02 平安科技（深圳）有限公司 Intelligent emotion recognition method and device and computer readable storage medium
CN111374651A (en) * 2020-04-10 2020-07-07 江苏米孚致远物联科技有限公司 Novel wisdom endowment system

Also Published As

Publication number Publication date
CN112183228A (en) 2021-01-05

Similar Documents

Publication Publication Date Title
CN112183228B (en) Intelligent community endowment service system and method
Chen et al. SPHA: Smart personal health advisor based on deep analytics
Sheeran et al. Can the theory of planned behavior explain patterns of health behavior change?
Yu et al. Personalized health monitoring system of elderly wellness at the community level in Hong Kong
Keren et al. End-to-end learning for dimensional emotion recognition from physiological signals
CN107680674A (en) A kind of healthy platform of internet of things and its management method based on healthy big data
Reyna Theories of medical decision making and health: an evidence-based approach
CN110570919A (en) Intelligent health risk monitoring and early warning method and platform for chronic disease people
CN111513732A (en) Intelligent psychological stress assessment early warning system for various groups of people under epidemic disease condition
CN110689935A (en) Internet of things emergency service platform and early warning method
CN116564561A (en) Intelligent voice nursing system and nursing method based on physiological and emotion characteristics
Deepika et al. Iot based elderly monitoring system
Dowrick Why do the O'Sheas consult so often? An exploration of complex family illness behaviour
Lopez et al. The RESTORE Program for sex crimes: differentiating therapeutic jurisprudence from restorative justice with therapeutic components
Guo et al. A personalized spatial-temporal cold pain intensity estimation model based on facial expression
CN112669927A (en) Household health service system
CN116098595B (en) System and method for monitoring and preventing sudden cardiac death and sudden cerebral death
CN116597954A (en) Intelligent home-based remote diagnosis and treatment appointment consultation system
Jayaraman et al. An ontology-based framework for real-time collection and visualization of mobile field triage data in mass gatherings
Orpa et al. An IoT based healthcare solution with ESP32 using machine learning model
CN113869179A (en) Multi-mode feature attention fused home-based and aged-care intelligent monitoring method and system
Orejel Bustos et al. Ambient assisted living systems for falls monitoring at home
Gurjar et al. Mental Health Prediction Using Machine Learning
Sun et al. An overview of aiomt applications
Fiala et al. Cues for facial attractiveness and preference of sexual dimorphism: A comparison across five cultures

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant