CN113892910A

CN113892910A - Internet of things-based intelligent closestool old people monitoring method

Info

Publication number: CN113892910A
Application number: CN202111141169.5A
Authority: CN
Inventors: 胡玮
Original assignee: Lichuan Xinwei Technology Co ltd
Current assignee: Lichuan Xinwei Technology Co ltd
Priority date: 2021-09-28
Filing date: 2021-09-28
Publication date: 2022-01-07

Abstract

The application relates to the field of Internet of things, and particularly discloses an old people monitoring method and system of an intelligent closestool based on the Internet of things and electronic equipment. The method adopts a convolutional neural network model based on a deep learning technology to excavate the relevance characteristics among various detection data of different old people, so that the detection of the physical condition of the old people is carried out based on the relevance characteristics among the detection data of a plurality of old people, and in the process, a converter model is further adopted to encode the relevance characteristics of implicit expression among the detection data of different old people to obtain the explicit expression of the relevance characteristics. By the method, the classification precision and accuracy are high, and the physical condition of the old can be better detected.

Description

Internet of things-based intelligent closestool old people monitoring method

Technical Field

The invention relates to the field of internet of things, in particular to an old people monitoring method and system of an intelligent closestool based on the internet of things and electronic equipment.

Background

With the increasingly prominent aging problem in China in recent years, the proportion of the aged population in the national population is continuously increased, so that the aged becomes a group which has to be paid attention to in all social circles. The physical and psychological changes of the elderly people begin to occur, and the design and the development of products for the elderly people, especially bathroom products, are short of demand.

The intelligence of household articles is accepted by people at the present that the Internet of things slowly enters our lives. Because the old person falls down or other unexpected circumstances appear when like the lavatory is the huge potential safety hazard of old person's solitary, consequently what adopt at present implants intelligent chip in the closestool, with the help of wireless network in order to realize the intellectuality of closestool, reaches the purpose of real time monitoring old person's healthy state, satisfies the requirement that old person was good at home and aged. However, most of the conventional detection methods monitor the physical condition of the elderly based on the detection data of a single elderly, for example, based on the single data of the elderly, such as blood pressure, weight, urine protein, blood protein content, and the like. Therefore, the relevance among the sample data can be ignored, and if one data in the sample data changes due to the influence of external environment factors, disturbance can be generated, so that the detection accuracy is low. Therefore, a scheme capable of intelligently monitoring whether the old people fall down when using the toilet is expected.

Disclosure of Invention

The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides an old people monitoring method, system and electronic equipment of an intelligent closestool based on the Internet of things, wherein a convolutional neural network model based on a deep learning technology is adopted to excavate relevance characteristics among various detection data of different old people, so that the physical condition of the old people is detected based on the relevance characteristics among the detection data of a plurality of old people, and in the process, a converter model is further adopted to code the relevance characteristics of implicit expression among the detection data of different old people so as to obtain explicit expression of the relevance characteristics. By the method, the classification precision and accuracy are high, and the physical condition of the old can be better detected.

According to an aspect of the application, a method for monitoring the old people of an intelligent closestool based on the internet of things is provided, and comprises the following steps:

the method comprises the steps that detection data of a plurality of old people to be detected are obtained through a plurality of intelligent toilets connected based on the Internet of things;

constructing a plurality of detection data of the elderly to be detected into an input matrix and obtaining a feature map through a convolutional neural network, wherein the feature map is used for representing implicit correlation among various detection data of different elderly;

matrix multiplication is carried out on the feature map and a plurality of detection vectors formed by a plurality of detection data of the old people to be detected respectively so as to map the detection vectors into feature spaces of the feature map respectively, and a plurality of feature vectors are obtained;

inputting the plurality of feature vectors into a converter model for context coding of each feature vector based on information of other feature vectors to obtain a plurality of coded feature vectors;

calculating a prediction classification value of each of the plurality of coding eigenvectors corresponding to the converter model to obtain a classification eigenvector composed of a plurality of prediction classification values, wherein the prediction classification value corresponding to the converter model is generated based on a matrix product, a relation factor and a distance between two associated coding eigenvectors of the plurality of coding eigenvectors, respectively;

inputting the classified feature vectors into a classifier to obtain a classification result, wherein the classification result is used for indicating whether the physical condition of the old to be detected is normal or not; and

and monitoring the old people to be detected based on the classification result.

In the above method for monitoring elderly people in an intelligent toilet based on the internet of things, constructing a plurality of detection data of elderly people to be detected as an input matrix and obtaining a characteristic diagram through a convolutional neural network, the method includes: arranging the detection data of each to-be-detected old person into a detection vector; arranging the detection vectors into the input matrix with the elderly as a sample dimension; and, the convolutional neural network obtaining the feature map from the input matrix in the following formula; the formula is:

f_i＝active(N_i×f_{i_1}+B_i)

wherein f is_{i_1}Is the input of the i-th convolutional neural network, f_iIs the output of the ith convolutional neural network, N_iIs the convolution kernel of the ith convolutional neural network, and B_iActive represents a nonlinear activation function for the bias vector of the ith layer of convolutional neural network.

In the foregoing method for monitoring elderly people in an intelligent toilet based on internet of things, calculating a predicted classification value of each of the plurality of coded feature vectors corresponding to the converter model to obtain a classification feature vector composed of a plurality of predicted classification values includes: and multiplying one of the plurality of coding feature vectors by the transpose of the other coding feature vector, adding a relation factor, and dividing by the square root of the distance between the coding feature vector and the other coding feature vector to obtain the prediction classification value of the coding feature vector.

In the above method for monitoring elderly people in an intelligent toilet based on the internet of things, the relationship factor represents whether there is a correlation between the coding feature vector and the other coding feature vector.

In the foregoing method for monitoring elderly people in an intelligent toilet based on internet of things, inputting the classification feature vector into a classifier to obtain a classification result, the method includes: inputting the classification characteristic vector into a class Softmax classification function to obtain class Softmax classification function values corresponding to characteristic values of all positions in the classification characteristic vector as probability values corresponding to whether the physical conditions of the old people to be detected are normal or not; and determining the classification result based on the comparison between the class Softmax classification function value and a preset threshold value.

In the foregoing method for monitoring the elderly based on an intelligent toilet of internet of things, based on the classification result, monitoring the elderly to be detected includes: and responding to the fact that at least one old person to be detected has abnormal physical condition in the classification result, and sending a warning signal to terminal equipment related to the corresponding old person to be detected.

According to another aspect of the application, a monitoring system for the elderly based on intelligent closestool of internet of things is provided, which includes:

the system comprises a detection data acquisition unit, a detection data processing unit and a detection data processing unit, wherein the detection data acquisition unit is used for acquiring detection data of a plurality of old people to be detected through a plurality of intelligent toilets connected based on the Internet of things;

the convolutional neural network processing unit is used for constructing the detection data of the elderly to be detected, which are obtained by the detection data obtaining units, into an input matrix and obtaining a feature map through a convolutional neural network, wherein the feature map is used for representing implicit correlation among various detection data of different elderly people;

the feature vector generating unit is used for respectively carrying out matrix multiplication on the feature map obtained by the convolutional neural network processing unit and a plurality of detection vectors formed by the detection data of the old people to be detected obtained by the detection data obtaining unit so as to respectively map the detection vectors into the feature space of the feature map to obtain a plurality of feature vectors;

a converter processing unit configured to input the plurality of feature vectors obtained by the feature vector generation unit into a converter model to obtain a plurality of encoded feature vectors, wherein the converter model is configured to perform context encoding on each feature vector based on information of other feature vectors;

a prediction classification value calculation unit configured to calculate a prediction classification value of the converter model for each of the plurality of encoded feature vectors obtained by the converter processing unit to obtain a classification feature vector composed of a plurality of prediction classification values, wherein the prediction classification values corresponding to the converter model are generated based on a matrix product, a relation factor, and a distance between two associated encoded feature vectors of the plurality of encoded feature vectors, respectively;

the classification unit is used for inputting the classification characteristic vector obtained by the prediction classification value calculation unit into a classifier to obtain a classification result, and the classification result is used for indicating whether the physical condition of the old to be detected is normal or not; and

and the monitoring unit is used for monitoring the old people to be detected based on the classification result obtained by the classification unit.

In the above-mentioned old person monitoring system based on intelligent closestool of thing networking, convolutional neural network processing unit includes: the detection vector arrangement subunit is used for arranging the detection data of each to-be-detected old person into a detection vector; an input matrix arrangement subunit configured to arrange the detection vectors obtained by the detection vector arrangement subunit into the input matrix with the old as a sample dimension; and a convolutional neural network subunit for obtaining the feature map from the input matrix obtained by the input matrix arrangement subunit by the convolutional neural network in the following formula; the formula is:

f_i＝active(N_i×f_{i_1}+B_i)

In the foregoing old people monitoring system of an intelligent toilet based on the internet of things, the prediction classification value calculating unit is further configured to: and multiplying one of the plurality of coding feature vectors by the transpose of the other coding feature vector, adding a relation factor, and dividing by the square root of the distance between the coding feature vector and the other coding feature vector to obtain the prediction classification value of the coding feature vector.

In the above-mentioned old people monitoring system of the intelligent toilet based on the internet of things, the relationship factor represents whether there is an association between the coding feature vector and the other coding feature vector.

In the above-mentioned old person monitoring system based on intelligent closestool of thing networking, the sorting unit includes: a probability value calculating operator unit, configured to input the classification feature vector into a class Softmax classification function to obtain a class Softmax classification function value corresponding to a feature value of each position in the classification feature vector, as a probability value corresponding to whether the physical condition of the elderly to be detected is normal; and the classification result determining subunit is used for determining the classification result based on the comparison between the class Softmax classification function value obtained by the probability value calculating subunit and a preset threshold value.

In the above-mentioned old person monitoring system based on intelligent closestool of thing networking, the monitoring unit is further used for: and responding to the fact that at least one old person to be detected has abnormal physical condition in the classification result, and sending a warning signal to terminal equipment related to the corresponding old person to be detected.

According to still another aspect of the present application, there is provided an electronic apparatus including: a processor; and a memory in which computer program instructions are stored, which, when executed by the processor, cause the processor to perform the method of monitoring elderly people in an intelligent toilet based on internet of things as described above.

According to yet another aspect of the present application, there is provided a computer readable medium having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the method for monitoring elderly people in an intelligent toilet based on internet of things as described above.

Compared with the prior art, the method, the system and the electronic equipment for monitoring the old people of the intelligent closestool based on the internet of things adopt the convolutional neural network model based on the deep learning technology to excavate the relevance characteristics among various detection data of different old people, so that the physical condition of the old people is detected based on the relevance characteristics among the detection data of a plurality of old people, and in the process, the converter model is further adopted to code the relevance characteristics of implicit expression among the detection data of different old people to obtain the explicit expression of the relevance characteristics. By the method, the classification precision and accuracy are high, and the physical condition of the old can be better detected.

Drawings

The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.

Fig. 1 is an application scene diagram of an old people monitoring method of an intelligent toilet based on the internet of things according to an embodiment of the application;

fig. 2 is a flowchart of an old people monitoring method of an intelligent toilet based on the internet of things according to an embodiment of the application;

fig. 3 is a schematic system architecture diagram of an old people monitoring method of an intelligent toilet based on the internet of things according to an embodiment of the application;

fig. 4 is a flowchart of constructing detection data of a plurality of elderly people to be detected as an input matrix and obtaining a characteristic diagram through a convolutional neural network in the elderly people monitoring method of the intelligent toilet based on the internet of things according to the embodiment of the application;

fig. 5 is a block diagram of an elderly people monitoring system for an intelligent toilet based on the internet of things according to an embodiment of the present application;

fig. 6 is a block diagram of a convolutional neural network processing unit in an elderly people monitoring system of an intelligent toilet based on the internet of things according to an embodiment of the present application;

fig. 7 is a block diagram of an electronic device according to an embodiment of the application.

Detailed Description

Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.

Overview of a scene

As mentioned above, the intelligence of the household articles has been accepted by people today when the internet of things slowly enters our lives. Because the old person falls down or other unexpected circumstances appear when like the lavatory is the huge potential safety hazard of old person's solitary, consequently what adopt at present implants intelligent chip in the closestool, with the help of wireless network in order to realize the intellectuality of closestool, reaches the purpose of real time monitoring old person's healthy state, satisfies the requirement that old person was good at home and aged. However, most of the conventional detection methods monitor the physical condition of the elderly based on the detection data of a single elderly, for example, based on the single data of the elderly, such as blood pressure, weight, urine protein, blood protein content, and the like. Therefore, the relevance among the sample data can be ignored, and if one data in the sample data changes due to the influence of external environment factors, disturbance can be generated, so that the detection accuracy is low. Therefore, a scheme capable of intelligently monitoring whether the old people fall down when using the toilet is expected.

Based on this, in the technical scheme of this application, through the intelligent closestool based on the thing networking, can acquire the detected data of a plurality of old people, consequently, replace in the traditional scheme based on the detected data of single old person comes the scheme that detects the health of old person, carry out the detection of old person's health based on the correlation characteristic between the detected data of a plurality of old people in the technical scheme of this application, that is, regard as a holistic sample set with the detected data of a plurality of old people, if certain sample in the sample set takes place to deviate, just can think that this sample has the anomaly, can avoid single sample because the disturbance that the change that various factors take place produces like this.

Specifically, a plurality of detection data of the elderly to be detected, including blood pressure, weight, urine protein, blood protein content, etc. of the elderly, are first obtained through a plurality of intelligent toilets based on the internet of things, and then the data are constructed into an input matrix, i.e., each row of the matrix corresponds to one elderly, and each column of the matrix corresponds to one item of detection data. Then, the input matrix is passed through a convolutional neural network model to obtain a feature map, so that correlation features among various detection data of different old people are mined.

At this time, the feature map may be multiplied by a detection vector formed by detection data of a certain elderly person to be detected, so that the detection vector is mapped into the feature space of the associated features to obtain a feature vector of the certain elderly person to be detected. If the characteristic vector is directly used for single classification whether the physical condition of the old people is normal or not through the classifier, because the associated characteristics of various detection data of different old people in the characteristic vector are implicitly expressed, the information is difficult to be utilized through the classifier, and the classification precision is low.

Based on this, in the technical solution of the present application, a transformer (transformer) model is further adopted to further encode the association features of the implicit expression, so as to obtain an explicit expression of the association features, that is, in a manner similar to the semantic understanding model, a certain word can contain context information through the association of the feature expressions between the words.

Specifically, the feature map is multiplied by a plurality of detection vectors formed by a plurality of obtained detection data of the elderly to be detected, so as to obtain a plurality of feature vectors, and then the plurality of feature vectors are input into the converter model, so that each feature vector is encoded based on information of other feature vectors, that is, context encoding is performed, so as to obtain a plurality of encoded feature vectors. And each coded feature vector also corresponds to an elderly person to be detected.

Further, using the predictive classification mechanism of the converter, for the classified coding feature vector, the predictive classification value is calculated, that is, the coding feature vector is multiplied by the transpose of another coding feature vector, then a relation factor is added, and then the square root of the distance between the coding feature vector and the another coding feature vector is divided, wherein the relation factor represents whether a relation exists between the coding feature vector and the another coding feature vector, which can be obtained based on the sample data of the corresponding elderly, for example, the elderly belonging to the same age group thinks that a relation exists, and the like.

Therefore, a prediction classification value is obtained for each coding feature vector except the classified coding feature vector in the plurality of coding feature vectors, then the obtained prediction classification values are arranged into the classification feature vectors, and a classification result of whether the physical condition of the old to be detected is normal or not can be obtained through a classifier. And by the method, classification results can be obtained for a plurality of even all the elderly to be detected at one time.

Based on this, the application provides an old person monitoring method of intelligent closestool based on thing networking, and it includes: the method comprises the steps that detection data of a plurality of old people to be detected are obtained through a plurality of intelligent toilets connected based on the Internet of things; constructing a plurality of detection data of the elderly to be detected into an input matrix and obtaining a feature map through a convolutional neural network, wherein the feature map is used for representing implicit correlation among various detection data of different elderly; matrix multiplication is carried out on the feature map and a plurality of detection vectors formed by a plurality of detection data of the old people to be detected respectively so as to map the detection vectors into feature spaces of the feature map respectively, and a plurality of feature vectors are obtained; inputting the plurality of feature vectors into a converter model for context coding of each feature vector based on information of other feature vectors to obtain a plurality of coded feature vectors; calculating a prediction classification value of each of the plurality of coding eigenvectors corresponding to the converter model to obtain a classification eigenvector composed of a plurality of prediction classification values, wherein the prediction classification value corresponding to the converter model is generated based on a matrix product, a relation factor and a distance between two associated coding eigenvectors of the plurality of coding eigenvectors, respectively; inputting the classified feature vectors into a classifier to obtain a classification result, wherein the classification result is used for indicating whether the physical condition of the old to be detected is normal or not; and monitoring the elderly to be detected based on the classification result.

Fig. 1 illustrates an application scenario diagram of an old people monitoring method of an intelligent toilet based on the internet of things according to an embodiment of the application. As shown in fig. 1, in this application scenario, first, a plurality of detection data of the elderly to be detected sitting on the intelligent toilet are obtained through a plurality of intelligent toilets based on the internet of things (e.g., T as illustrated in fig. 1), wherein the detection data includes, but is not limited to, blood pressure, weight, urine protein, blood protein content, etc. of the elderly. Then, the acquired detection data of the plurality of elderly people to be detected is input into a server (for example, S shown in fig. 1) deployed with an intelligent toilet based on the internet of things for an elderly people monitoring algorithm, where the server can process the acquired detection data of the plurality of elderly people to be detected by using the intelligent toilet based on the internet of things for the elderly people monitoring algorithm to generate a classification result indicating whether the physical condition of the elderly people to be detected is normal, so that the elderly people to be detected can be monitored based on the classification result.

Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.

Exemplary method

Fig. 2 illustrates a flowchart of an old people monitoring method of an internet of things based intelligent toilet. As shown in fig. 2, the method for monitoring the elderly people of the intelligent toilet based on the internet of things according to the embodiment of the application includes: s110, obtaining detection data of a plurality of old people to be detected through a plurality of intelligent toilets connected based on the Internet of things; s120, constructing a plurality of detection data of the old people to be detected into an input matrix and obtaining a feature map through a convolutional neural network, wherein the feature map is used for representing implicit correlation among various detection data of different old people; s130, matrix multiplication is carried out on the feature map and a plurality of detection vectors formed by a plurality of detection data of the old people to be detected respectively, and the detection vectors are mapped into feature spaces of the feature map respectively to obtain a plurality of feature vectors; s140, inputting the plurality of feature vectors into a converter model to obtain a plurality of coded feature vectors, wherein the converter model is used for carrying out context coding on each feature vector based on information of other feature vectors; s150, calculating a prediction classification value of each of the plurality of coding feature vectors corresponding to the converter model to obtain a classification feature vector consisting of a plurality of prediction classification values, wherein the prediction classification values corresponding to the converter model are generated based on a matrix product, a relation factor and a distance between two associated coding feature vectors in the plurality of coding feature vectors respectively; s160, inputting the classification feature vectors into a classifier to obtain a classification result, wherein the classification result is used for indicating whether the physical condition of the old to be detected is normal or not; and S170, monitoring the elderly to be detected based on the classification result.

Fig. 3 illustrates an architecture diagram of an elderly people monitoring method of an intelligent toilet based on the internet of things according to an embodiment of the application. As shown IN fig. 3, IN the network architecture of the method for monitoring elderly people IN an intelligent toilet based on internet of things, firstly, the acquired detection data (e.g., IN1 as illustrated IN fig. 3) of the plurality of elderly people to be detected are constructed into an input matrix (e.g., M1 as illustrated IN fig. 3) and passed through a convolutional neural network (e.g., CNN1 as illustrated IN fig. 3) to obtain a feature map (e.g., F1 as illustrated IN fig. 3); s130, matrix-multiplying the feature map with a plurality of detection vectors (e.g., V1 as illustrated in fig. 3) composed of a plurality of detection data of the elderly to be detected, respectively, to map the detection vectors into feature spaces of the feature map, respectively, to obtain a plurality of feature vectors (e.g., VF1 as illustrated in fig. 3); s140, inputting the plurality of feature vectors into a converter model (e.g., D as illustrated in fig. 3) to obtain a plurality of encoded feature vectors (e.g., VF2 as illustrated in fig. 3); s150, calculating a predicted classification value of the converter model corresponding to each of the plurality of encoded feature vectors to obtain a classification feature vector (e.g., VF3 as illustrated in fig. 3) composed of a plurality of predicted classification values; s160, inputting the classification feature vector into a classifier (e.g. circle S as illustrated in fig. 3) to obtain a classification result, wherein the classification result is used for indicating whether the physical condition of the elderly to be detected is normal; and S170, monitoring the elderly to be detected based on the classification result.

In step S110, a plurality of detection data of the elderly to be detected are obtained through a plurality of intelligent toilets connected based on the internet of things. As described above, in the conventional scheme, the body condition of the elderly is detected based on the detection data of a single elderly person, and in the present application, the detection of the body condition of the elderly is performed by using the correlation characteristics between the detection data of a plurality of elderly persons, that is, the detection data of a plurality of elderly persons are regarded as a whole sample set, and if a certain sample in the sample set deviates, the sample can be considered to be abnormal, so that the disturbance of the single sample due to the change of various factors can be avoided. Therefore, in the technical scheme of this application, just at first need obtain a plurality of detection data that wait to detect the old person through a plurality of intelligent closestool that link to each other based on the thing networking. In a specific example, the elderly person to be detected may sit on the intelligent toilet, and the detection data includes, but is not limited to, blood pressure, weight, urine protein, blood protein content, and the like of the elderly person.

In step S120, a plurality of detection data of the elderly to be detected are constructed as an input matrix and passed through a convolutional neural network to obtain a feature map, wherein the feature map is used for representing implicit relations between the detection data of different elderly. It should be understood that the present application detects the physical condition of the elderly based on the correlation characteristics between the detected data of a plurality of elderly people, so as to avoid the disturbance influence of the change of a single sample due to various factors. Therefore, in the technical solution of the present application, firstly, the detection data of the plurality of elderly people to be detected are constructed as an input matrix, each row of the input matrix corresponds to one elderly person, and each column of the input matrix corresponds to one item of detection data. Then, the input matrix passes through a convolutional neural network model, and is processed through a convolutional neural network so as to mine the associated features among various detection data of different old people, and therefore a feature map is obtained. It is worth mentioning that here the feature map represents an implicit correlation between the various test data items of different elderly persons.

Specifically, in the embodiment of the present application, a process of constructing a plurality of detection data of an elderly person to be detected as an input matrix and obtaining a feature map by a convolutional neural network includes: firstly, arranging the detection data of each to-be-detected old person into a detection vector; then, arranging the detection vectors into the input matrix by taking the old as a sample dimension; finally, the convolutional neural network obtains the feature map from the input matrix in the following formula; the formula is:

f_i＝active(N_i×f_{i_1}+B_i)

Fig. 4 illustrates a flowchart of constructing detection data of a plurality of elderly people to be detected as an input matrix and obtaining a characteristic diagram through a convolutional neural network in the elderly people monitoring method of the intelligent toilet based on the internet of things according to the embodiment of the application. As shown in fig. 4, constructing a plurality of detection data of the elderly to be detected as an input matrix and obtaining a feature map by convolving a neural network, includes: s210, arranging the detection data of each to-be-detected old person into a detection vector; s220, arranging the detection vectors into the input matrix by taking the old as a sample dimension; and S230, passing the input matrix through a convolutional neural network to obtain a feature map.

In steps S130 and S140, matrix multiplication is performed on the feature map and a plurality of detection vectors composed of a plurality of detection data of the elderly to be detected respectively to map the detection vectors into feature spaces of the feature map respectively to obtain a plurality of feature vectors, and the plurality of feature vectors are input into a converter model to obtain a plurality of coded feature vectors, wherein the converter model is used for context coding of information based on other feature vectors for each feature vector. It should be understood that after the feature map is obtained, a detection vector formed by multiplying the feature map by detection data of a certain elderly person to be detected may be used to map the detection vector into a feature space of associated features, so as to obtain a feature vector of the certain elderly person to be detected. However, if the feature vector is directly used to perform a single classification of whether the physical condition of the elderly is normal or not through a classifier, since the associated features between the items of detection data about different elderly in the feature vector are implicitly expressed, it is difficult to use this information through the classifier, and thus the classification accuracy is not high. Therefore, in the technical solution of the present application, a converter model is further adopted to further encode the association features of implicit expression, so as to obtain explicit expression of the association features, that is, in a manner similar to a semantic understanding model, a word can contain context information through association of feature expression between words.

Specifically, first, a plurality of detection vectors formed by multiplying the feature map by the obtained detection data of the plurality of elderly people to be detected are multiplied, that is, the plurality of detection vectors are mapped into the feature space of the feature map, so as to obtain a plurality of feature vectors; the plurality of feature vectors are then input into a converter model such that each feature vector is encoded, i.e. context encoded, based on information of other feature vectors to obtain a plurality of encoded feature vectors. It is worth mentioning that here each of said coded feature vectors also corresponds to an elderly person to be detected.

In step S150, a prediction classification value of each of the plurality of coding feature vectors corresponding to the converter model is calculated to obtain a classification feature vector composed of a plurality of prediction classification values, wherein the prediction classification value corresponding to the converter model is generated based on a matrix product, a relation factor and a distance between two associated coding feature vectors of the plurality of coding feature vectors, respectively. It should be understood that in order to make the classifier use the implicit correlation features between the various items of detected data of different elderly people to make the classification result more accurate and precise, it is necessary to calculate the predicted classification value of the converter model for each of the plurality of encoded feature vectors. That is, using the predictive classification mechanism of the converter, for the classified encoded eigenvector, the predictive classification value is calculated, i.e., the encoded eigenvector is multiplied by the transpose of another encoded eigenvector, then the relationship factor is added, and then the square root of the distance between the encoded eigenvector and the other encoded eigenvector is divided. It is worth mentioning that, here, the relationship factor represents whether there is a relationship between the encoding feature vector and the other encoding feature vector, and in a specific example, this may be obtained based on sample data of an elderly person corresponding to the relationship, for example, elderly persons belonging to the same age group think there is a relationship, and the like. In this way, for each of the plurality of encoded feature vectors other than the classified encoded feature vector, a prediction classification value is obtained, and then, the obtained prediction classification values can be arranged as a classification feature vector.

Specifically, in the embodiment of the present application, the process of calculating the prediction classification value of the converter model corresponding to each of the plurality of coding feature vectors to obtain a classification feature vector composed of a plurality of prediction classification values includes: and multiplying one of the plurality of coding feature vectors by the transpose of the other coding feature vector, adding a relation factor, and dividing by the square root of the distance between the coding feature vector and the other coding feature vector to obtain the prediction classification value of the coding feature vector. It is worth mentioning that here, the relationship factor indicates whether there is an association between the encoded feature vector and the other encoded feature vector.

In step S160 and step S170, the classification feature vector is input into a classifier to obtain a classification result, the classification result is used for indicating whether the physical condition of the elderly to be detected is normal, and the elderly to be detected is monitored based on the classification result. In a specific example, firstly, the classification feature vector is input into a class Softmax classification function to obtain class Softmax classification function values corresponding to feature values of various positions in the classification feature vector as probability values corresponding to whether the physical conditions of the elderly to be detected are normal or not. Then, based on the comparison between the class Softmax classification function value and a preset threshold value, the classification result is determined. Specifically, when the class Softmax classification function value is larger than the preset threshold value, the classification result indicates that the body condition of the corresponding elderly to be detected is abnormal; and when the class Softmax classification function value is smaller than the preset threshold value, the classification result indicates that the corresponding physical condition of the old to be detected is normal. And finally, in response to the fact that at least one old person to be detected has abnormal physical condition in the classification result, sending a warning signal to terminal equipment related to the corresponding old person to be detected.

In summary, the method for monitoring the elderly people of the intelligent toilet based on the internet of things according to the embodiment of the present application is elucidated, the convolutional neural network model based on the deep learning technology is adopted to dig out the relevance characteristics among various detection data of different elderly people, so that the physical condition of the elderly people is detected based on the relevance characteristics among the detection data of a plurality of elderly people, and in the process, the converter model is further adopted to code the relevance characteristics of the implicit expression among the detection data of different elderly people to obtain the explicit expression of the relevance characteristics. By the method, the classification precision and accuracy are high, and the physical condition of the old can be better detected.

Exemplary System

Fig. 5 illustrates a block diagram of an elderly people monitoring system of an intelligent toilet based on internet of things according to an embodiment of the present application. As shown in fig. 5, an old people monitoring system 500 of an intelligent toilet based on internet of things according to an embodiment of the present application includes: a detection data acquisition unit 510, configured to acquire detection data of a plurality of elderly people to be detected through a plurality of intelligent toilets connected based on the internet of things; a convolutional neural network processing unit 520, configured to construct the detection data of the elderly to be detected, obtained by the multiple detection data obtaining units 510, into an input matrix and obtain a feature map through a convolutional neural network, where the feature map is used to represent implicit associations between various detection data of different elderly people; a feature vector generating unit 530, configured to perform matrix multiplication on the feature map obtained by the convolutional neural network processing unit 520 and a plurality of detection vectors composed of the detection data of the plurality of elderly people to be detected obtained by the detection data obtaining unit 510, so as to map the detection vectors into feature spaces of the feature map, respectively, so as to obtain a plurality of feature vectors; a converter processing unit 540, configured to input the plurality of feature vectors obtained by the feature vector generation unit 530 into a converter model to obtain a plurality of encoded feature vectors, wherein the converter model is used for context encoding of each feature vector based on information of other feature vectors; a prediction classification value calculating unit 550, configured to calculate a prediction classification value of the converter model for each of the plurality of coding feature vectors obtained by the converter processing unit 540, so as to obtain a classification feature vector composed of a plurality of prediction classification values, where the prediction classification values corresponding to the converter model are generated based on a matrix product, a relation factor, and a distance between two associated coding feature vectors of the plurality of coding feature vectors, respectively; a classification unit 560, configured to input the classification feature vector obtained by the prediction classification value calculation unit 550 into a classifier to obtain a classification result, where the classification result is used to indicate whether the physical condition of the elderly to be detected is normal; and a monitoring unit 570, configured to monitor the elderly to be detected based on the classification result obtained by the classification unit 560.

In an example, in the above-mentioned system 500 for monitoring elderly people with intelligent toilet based on internet of things, as shown in fig. 6, the convolutional neural network processing unit 520 includes: a detection vector arrangement subunit 521, configured to arrange the detection data of each of the elderly people to be detected into a detection vector; an input matrix arrangement subunit 522 configured to arrange the detection vectors obtained by the detection vector arrangement subunit 521 into the input matrix with the elderly as a sample dimension; and a convolutional neural network subunit 523 configured to obtain the feature map from the input matrix obtained by the input matrix arrangement subunit 522 by the following formula for the convolutional neural network; the formula is:

f_i＝active(N_i×f_{i_1}+B_i)

wherein f is_{i_1}For i-th convolutional neural networksInput, f_iIs the output of the ith convolutional neural network, N_iIs the convolution kernel of the ith convolutional neural network, and B_iActive represents a nonlinear activation function for the bias vector of the ith layer of convolutional neural network.

In an example, in the above-mentioned system 500 for monitoring elderly people with intelligent toilet based on internet of things, the unit 550 for calculating a predicted classification value is further configured to: and multiplying one of the plurality of coding feature vectors by the transpose of the other coding feature vector, adding a relation factor, and dividing by the square root of the distance between the coding feature vector and the other coding feature vector to obtain the prediction classification value of the coding feature vector.

In one example, in the above-mentioned system 500 for monitoring elderly people with intelligent toilets based on internet of things, the relationship factor represents whether there is an association between the coded feature vector and the other coded feature vector.

In one example, in the above-mentioned system 500 for monitoring elderly people with intelligent toilet based on internet of things, the classification unit 560 includes: a probability value calculating operator unit, configured to input the classification feature vector into a class Softmax classification function to obtain a class Softmax classification function value corresponding to a feature value of each position in the classification feature vector, as a probability value corresponding to whether the physical condition of the elderly to be detected is normal; and the classification result determining subunit is used for determining the classification result based on the comparison between the class Softmax classification function value obtained by the probability value calculating subunit and a preset threshold value.

In one example, in the foregoing system 500 for monitoring elderly people with intelligent toilets based on internet of things, the monitoring unit 570 is further configured to: and responding to the fact that at least one old person to be detected has abnormal physical condition in the classification result, and sending a warning signal to terminal equipment related to the corresponding old person to be detected.

Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the above-described internet-of-things-based intelligent toilet elderly monitoring system 500 have been described in detail in the above description of the internet-of-things-based intelligent toilet elderly monitoring method with reference to fig. 1 to 4, and thus, a repetitive description thereof will be omitted.

As described above, the system 500 for monitoring the elderly people in the intelligent toilet based on the internet of things according to the embodiment of the present application can be implemented in various terminal devices, for example, a server of an algorithm for monitoring the elderly people in the intelligent toilet based on the internet of things, and the like. In one example, the system 500 for monitoring the elderly people based on intelligent closestool of internet of things according to the embodiment of the present application can be integrated into a terminal device as a software module and/or a hardware module. For example, the system 500 for monitoring elderly people based on intelligent toilets of internet of things may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the system 500 for monitoring the elderly people based on the intelligent toilet of internet of things can also be one of numerous hardware modules of the terminal device.

Alternatively, in another example, the system for monitoring elderly people with intelligent toilet based on internet of things 500 and the terminal device may be separate devices, and the system for monitoring elderly people with intelligent toilet based on internet of things 500 may be connected to the terminal device through a wired and/or wireless network and transmit the interaction information according to the agreed data format.

Exemplary electronic device

Next, an electronic apparatus according to an embodiment of the present application is described with reference to fig. 7. As shown in fig. 7, the electronic device 10 includes one or more processors 11 and a memory 12. The processor 11 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.

Memory 12 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, Random Access Memory (RAM), cache memory (cache), and/or the like. The non-volatile memory may include, for example, Read Only Memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer readable storage medium and executed by the processor 11 to implement the functions of the method for monitoring elderly people in an intelligent toilet based on internet of things of the various embodiments of the present application described above and/or other desired functions. Various content such as coded feature vectors, classified feature vectors, and the like may also be stored in the computer-readable storage medium.

In one example, the electronic device 10 may further include: an input system 13 and an output system 14, which are interconnected by a bus system and/or other form of connection mechanism (not shown).

The input system 13 may comprise, for example, a keyboard, a mouse, etc.

The output system 14 may output various information including classification results and the like to the outside. The output system 14 may include, for example, a display, speakers, a printer, and a communication network and its connected remote output devices, among others.

Of course, for simplicity, only some of the components of the electronic device 10 relevant to the present application are shown in fig. 7, and components such as buses, input/output interfaces, and the like are omitted. In addition, the electronic device 10 may include any other suitable components depending on the particular application.

Exemplary computer program product and computer-readable storage Medium

In addition to the above-described methods and apparatus, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the functions of the method for monitoring for the elderly with intelligent toilets based on internet of things according to various embodiments of the present application described in the "exemplary methods" section of this specification above.

The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.

Furthermore, embodiments of the present application may also be a computer readable storage medium having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the steps in the method for monitoring elderly people with an intelligent toilet based on internet of things described in the "exemplary methods" section above in this specification.

The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.

The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".

It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.

The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.

Claims

1. The monitoring method for the old people of the intelligent closestool based on the Internet of things is characterized by comprising the following steps:

2. The method for monitoring the elderly people in the intelligent toilet based on the internet of things according to claim 1, wherein constructing the detection data of a plurality of elderly people to be detected as an input matrix and obtaining the characteristic diagram through a convolutional neural network comprises:

arranging the detection data of each to-be-detected old person into a detection vector;

arranging the detection vectors into the input matrix with the elderly as a sample dimension; and

the convolutional neural network obtains the feature map from the input matrix in the following formula;

the formula is:

f_i＝active(N_i×f_i-1+B_i)

wherein f is_i-1Is the input of the i-th convolutional neural network, f_iIs the output of the ith convolutional neural network, N_iIs the convolution kernel of the ith convolutional neural network, and B_iActive represents a nonlinear activation function for the bias vector of the ith layer of convolutional neural network.

3. The method for monitoring the elderly people in an intelligent toilet based on internet of things according to claim 1, wherein calculating a predicted classification value of the converter model corresponding to each of the plurality of coded feature vectors to obtain a classification feature vector composed of a plurality of predicted classification values comprises:

and multiplying one of the plurality of coding feature vectors by the transpose of the other coding feature vector, adding a relation factor, and dividing by the square root of the distance between the coding feature vector and the other coding feature vector to obtain the prediction classification value of the coding feature vector.

4. The Internet of things-based intelligent closestool old people monitoring method according to claim 3, wherein the relation factor represents whether an association exists between the coding feature vector and the other coding feature vector.

5. The method for monitoring the elderly people in an intelligent toilet based on the internet of things according to claim 1, wherein the inputting the classification feature vector into a classifier to obtain a classification result comprises:

inputting the classification characteristic vector into a class Softmax classification function to obtain class Softmax classification function values corresponding to characteristic values of all positions in the classification characteristic vector as probability values corresponding to whether the physical conditions of the old people to be detected are normal or not; and

determining the classification result based on a comparison between the class Softmax classification function value and a preset threshold.

6. The method for monitoring the elderly people of the intelligent closestool based on the Internet of things according to claim 1, wherein the monitoring of the elderly people to be detected based on the classification result comprises:

and responding to the fact that at least one old person to be detected has abnormal physical condition in the classification result, and sending a warning signal to terminal equipment related to the corresponding old person to be detected.

7. The utility model provides an old person monitoring system of intelligent closestool based on thing networking which characterized in that includes:

8. The intelligent closestool based on the internet of things old people monitoring system as claimed in claim 7, wherein the convolutional neural network processing unit comprises:

the detection vector arrangement subunit is used for arranging the detection data of each to-be-detected old person into a detection vector;

an input matrix arrangement subunit configured to arrange the detection vectors obtained by the detection vector arrangement subunit into the input matrix with the old as a sample dimension; and

a convolutional neural network subunit for the convolutional neural network to obtain the feature map from the input matrix obtained by the input matrix arrangement subunit in the following formula;

the formula is:

f_i＝active(N_i×f_i-1+B_i)

9. The internet of things based intelligent toilet elderly monitoring system of claim 7, wherein the predictive classification value calculating unit is further configured to:

10. An electronic device, comprising:

a processor; and

a memory having stored therein computer program instructions which, when executed by the processor, cause the processor to perform the method of monitoring for elderly people in an internet of things based smart toilet of any of claims 1-6.