Intelligent old-age care digital management system and method based on cloud platform

Abstract

The invention discloses an intelligent old-age care digital management system and an intelligent old-age care digital management method based on a cloud platform. The acquisition module is used for acquiring basic health data information of the old and sending the information to the cloud platform; the path comparison module is used for acquiring real-time positions of the old people, generating real-time paths of the old people according to the real-time positions, and performing similarity calculation on the real-time paths and historical paths to obtain path similarity; the emotion recognition module is used for recognizing the emotional state of the elderly and giving an emotional state label; the health state recognition module obtains real-time health data information of the elderly, constructs a health state recognition model, trains the health state recognition model to obtain a trained health state recognition model, and recognizes the health state of the elderly to obtain a recognition result; the cloud platform judges the behavior of the old person according to the identification result, and performs early warning if the behavior is abnormal; and under the unattended condition, management personnel or family members can quickly find the abnormal state of the old people and process the abnormal state in time.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of old-age management, in particular to a smart old-age digital management system and method based on a cloud platform. BACKGROUND

[0002] With the intensification of the population aging trend, the old-age dependency ratio is rising, and the traditional old-age service mode is facing many challenges, such as rising labor costs, uneven service quality, low management efficiency, etc. Old-age digital management system is rapidly developed. The old-age digital management system is suitable for old-age homes, homes for the aged, nursing homes, old-age welfare homes and other institutions to improve service quality and efficiency.

[0003] However, the existing old-age management system lacks the management function of the safety monitoring of the old people, and when there is no one to take care of, if the old people have an accident, it cannot be discovered and warned in time, which is easy to cause the old people to miss the rescue time. SUMMARY

[0004] In view of the above prior art, the present application provides a smart old-age digital management system and method based on a cloud platform, which mainly solves the technical problems existing in the background art.

[0005] To achieve the above purpose, the technical scheme of the embodiment of the present application is as follows:

[0006] In a first aspect, the present application provides a smart old-age digital management system based on a cloud platform, comprising:

[0007] An acquisition module is configured to acquire basic health data information of the old people and send the basic health data information to the cloud platform;

[0008] A path comparison module is configured to acquire a real-time position of the old people, generate a real-time path of the old people according to the real-time position, and perform similarity calculation on the real-time path and a historical path to obtain a path similarity;

[0009] An emotion recognition module is configured to recognize the emotional state of the old people and assign an emotional state label;

[0010] A health state recognition module is configured to acquire real-time health data information of the old people, construct a health state recognition model, input the basic health data information, the real-time health data information, the path similarity and the emotional state label into the health state recognition model for training, obtain a trained health state recognition model, and recognize the health state of the old people according to the trained health state recognition model to obtain a recognition result;

[0011] The cloud platform is configured to judge whether the old people have abnormal behavior according to the recognition result, and if so, to give a warning.

[0012] Optionally, the path comparison module includes a first identifier, a location detection unit, and a path calculation unit;

[0013] The first identifier stores the elderly person's basic information, and the first identifier is an active FRID tag;

[0014] The location detection unit includes multiple units, which are evenly distributed in the activity area. The distance between each location detection unit and the first marker is calculated, and the distance is sent to the cloud platform. The cloud platform generates the elderly person's real-time path in chronological order.

[0015] The path calculation unit is used to obtain the elderly person's historical path from the cloud platform, and calculate the similarity between the historical path and the real-time path based on Euclidean distance to obtain the path similarity.

[0016] Optionally, the process of identifying the elderly person's emotional state and assigning an emotional state label includes:

[0017] Collect video stream data captured by real-time monitoring cameras in the area where the elderly person is located, and divide the video stream data into frames to obtain image data;

[0018] The image data is processed using deep learning detection boxes to obtain the facial recognition results in each frame of the image, and the feature reference points of facial expressions are determined based on the facial recognition results.

[0019] A facial feature mesh is generated based on the feature reference points. Deformation is determined based on the facial feature mesh. The similarity between the deformation and the reference emotional feature mesh is calculated. The emotional state of the elderly person is determined based on the similarity calculation result, and an emotional state label is assigned.

[0020] Optionally, the health data includes heart rate, blood pressure, height, and weight;

[0021] Acquire real-time health data of the elderly, construct a health status recognition model, and input the basic health data, the real-time health data, the path similarity, and the emotional state label into the health status recognition model for training to obtain a trained health status recognition model, including:

[0022] The system acquires real-time health data of the elderly, calculates the rate of change of health data based on the real-time health data and the basic health data, and integrates the rate of change of health data with the real-time health data in chronological order to obtain a first health data feature set.

[0023] The first health data feature set is input into XGboot for feature filtering to obtain the second health data feature set;

[0024] A health status recognition model is constructed based on LSTM. The second health data feature set, the emotional state label, and the path similarity are input into the health status recognition model for training to obtain a trained health status recognition model.

[0025] Optionally, it also includes a second identifier, which is bound to the mobile terminal of the administrator or the elderly person's family member, and the cloud platform stores the mapping relationship between the second identifier and the mobile terminal of the administrator;

[0026] If abnormal behavior is detected, an early warning will be issued, including:

[0027] The cloud platform sends the location of the first identifier to the mobile terminal bound to the second identifier to issue an early warning.

[0028] Optionally, it also includes an electricity monitoring module, which is used to acquire the electricity consumption information of the indoor equipment when the elderly are in the indoor area, send the electricity consumption information to the cloud platform, the cloud platform analyzes the electricity consumption information, obtains the electricity consumption analysis results, judges whether the elderly's electricity consumption information is reasonable based on the electricity consumption analysis results, generates an electricity consumption message, and sends the electricity consumption message to the mobile terminal of the management personnel.

[0029] Optionally, the step of acquiring equipment power consumption information data in the indoor area, sending the power consumption information data to a cloud platform, the cloud platform performing power consumption analysis on the power consumption information data to obtain power consumption analysis results, determining whether the elderly person's power consumption information is reasonable based on the power consumption analysis results, generating a power consumption message, and sending the power consumption message to the administrator's mobile terminal includes:

[0030] Set a period for monitoring the elderly's electricity consumption, obtain electricity consumption information data during the monitoring period, and send the electricity consumption information data to the cloud platform;

[0031] The cloud platform stores preset electricity consumption thresholds for the elderly during the electricity monitoring period. The electricity consumption thresholds are compared with the electricity consumption information data. If the electricity consumption information data is greater than the electricity consumption thresholds, an electricity consumption growth rate analysis is performed.

[0032] If the electricity consumption data is less than the electricity consumption threshold, a normal electricity consumption message is generated and sent to the administrator's mobile terminal.

[0033] Optionally, the electricity consumption growth rate analysis includes:

[0034] The power consumption data of each electrical device connected to the smart socket during the current power consumption monitoring period is obtained through the smart socket, and the power consumption data during the current power consumption monitoring period is sent to the cloud platform.

[0035] The cloud platform obtains the electricity consumption data from the previous electricity consumption monitoring period, calculates the difference between the electricity consumption data from the current electricity consumption monitoring period and the electricity consumption data from the previous electricity consumption monitoring period, and obtains the electricity consumption amplitude of each electrical device. The cloud platform presets an electricity consumption amplitude threshold, compares the electricity consumption amplitude with the electricity consumption amplitude threshold, and if the electricity consumption amplitude is greater than the electricity consumption amplitude threshold, generates an abnormal electricity consumption growth rate message and sends the abnormal electricity consumption growth rate message to the mobile terminal of the management personnel.

[0036] If the electricity consumption amplitude is less than the electricity consumption amplitude threshold, a normal electricity consumption growth rate message is generated and sent to the mobile terminal of the management personnel.

[0037] Secondly, the present invention also provides a cloud-based smart elderly care digital management method, comprising the following steps:

[0038] Obtain basic health data information of the elderly and send the basic health data information to the cloud platform for storage;

[0039] Obtain the elderly person's real-time location information, generate a real-time path based on the real-time location information, calculate the similarity between the real-time path and the historical path, and obtain the path similarity.

[0040] Facial recognition is performed on frame sequence images of surveillance video based on deep learning detection boxes. Based on the facial recognition results, the emotional state of the elderly is identified, and an emotional state label is assigned.

[0041] The elderly person's real-time health data is obtained, a health status recognition model is constructed, and the path similarity, emotional state label, and real-time health data are input into the health status recognition model for training to obtain the recognition result.

[0042] Based on the identification results, it is determined whether the elderly person exhibits abnormal behavior. If abnormal behavior is found, an early warning is issued.

[0043] The beneficial effects of this invention are as follows: This invention provides a cloud-based smart elderly care digital management system. It acquires basic health data of the elderly through an acquisition module and stores it in the cloud platform. By calculating the similarity between the elderly's current path and historical paths, a path similarity score is obtained. Furthermore, by identifying the elderly's emotions, an emotional state label is obtained. The path similarity score, emotional state label, basic health data, and real-time health data are input into a health status recognition model to identify the elderly's state and obtain the recognition result. If an abnormal recognition result is found, it is sent to management personnel or family members, enabling them to quickly understand the elderly's abnormal state and take timely action. This allows for rapid detection of elderly abnormalities even when unattended, preventing missed opportunities for rescue. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the structure of a cloud-based smart elderly care digital management system provided in an embodiment of the present invention;

[0045] Figure 2 This is a flowchart illustrating a cloud-based smart elderly care digital management method provided in an embodiment of the present invention. Detailed Implementation

[0046] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. In the following description, the expression "some embodiments" refers to a subset of all possible embodiments; however, it should be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with each other without conflict.

[0047] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0048] It should be understood that the present invention can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein. Rather, providing these embodiments will make the disclosure thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Furthermore, the terminology used herein is intended only to describe particular embodiments and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms “compose” and / or “comprising,” when used in this specification, identify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.

[0049] It should also be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "inner," "outer," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0050] To fully understand this invention, a detailed structure will be presented in the following description to illustrate the technical solution proposed by this invention. Optional embodiments of the invention are described in detail below; however, in addition to these detailed descriptions, the invention may have other embodiments.

[0051] Example 1

[0052] Please refer to the attached document. Figure 1 This application provides a cloud-based smart elderly care digital management system, comprising:

[0053] The acquisition module is used to acquire the basic health data information of the elderly and send the basic health data information to the cloud platform;

[0054] The path comparison module is used to obtain the elderly person's real-time location, generate the elderly person's real-time path based on the real-time location, and calculate the similarity between the real-time path and the historical path to obtain the path similarity.

[0055] The emotion recognition module is used to identify the emotional state of the elderly and assign emotional state labels.

[0056] The health status recognition module acquires the elderly's real-time health data information, constructs a health status recognition model, and inputs the basic health data information, the real-time health data information, the path similarity, and the emotional state label into the health status recognition model for training to obtain a trained health status recognition model. The elderly's health status is then recognized based on the trained health status recognition model to obtain the recognition result.

[0057] The cloud platform determines whether the elderly person exhibits abnormal behavior based on the identification results, and issues an alert if abnormal behavior is detected.

[0058] Specifically, the module acquires basic health data of the elderly and stores it in the cloud platform. This effectively preserves the elderly's basic health data and provides insight into their basic health status, facilitating subsequent comparison with real-time health data to assess their overall health. In nursing homes and other elderly care facilities, the elderly's activity range, time, and routes are relatively fixed. Therefore, in outdoor areas, the path comparison module generates the elderly's real-time path and calculates its similarity to their historical paths. This quantifies the degree of deviation between the real-time and historical paths, thereby determining whether the elderly's behavior has changed. For example, if the elderly's behavior at a certain location differs from their historical path... If an elderly person spends too much time on a particular path, they may fall or faint, or their path may change significantly, potentially indicating abnormal behavior. Next, their emotional state is assessed and labeled. Then, real-time health data is acquired, and the rate of change in health data is calculated based on this data and baseline data. This rate of change, along with the emotional state label and path similarity, is input into the health status recognition model to identify the elderly person's health status. By comprehensively considering location, emotion, and health status, the model can better identify the elderly person's condition and more quickly detect abnormal behavior, facilitating appropriate interventions by management.

[0059] As an optional implementation, the path comparison module includes a first identifier, a location detection unit, and a path calculation unit;

[0060] The first identifier stores the elderly person's basic information, and the first identifier is an active FRID tag;

[0061] The location detection unit includes multiple units, which are evenly distributed in the activity area. The distance between each location detection unit and the first marker is calculated, and the distance is sent to the cloud platform. The cloud platform generates the elderly person's real-time path in chronological order.

[0062] The path calculation unit is used to obtain the elderly person's historical path from the cloud platform, and calculate the similarity between the historical path and the real-time path based on Euclidean distance to obtain the path similarity.

[0063] Specifically, the first identifier stores the elderly person's basic information and is bound to the elderly person's mobile terminal. For example, the mobile terminal can be a mobile phone or a wristband. The first identifier is an active RFID tag and can communicate with the cloud platform wirelessly. The location detection unit detects the location of the first identifier and the distance between itself and the first identifier, and sends the distance data to the cloud platform. The cloud platform generates the elderly person's real-time path based on the time sequence of the multiple detection units received. It also obtains the elderly person's historical path from the cloud platform, calculates the Euclidean distance between the historical path and the real-time path, and determines the similarity between the historical path and the real-time path through the Euclidean distance to obtain the path similarity.

[0064] The higher the similarity, the more similar the historical path and the real-time path are, and the smaller the deviation of the real-time path from the historical path, indicating that the elderly person's path has not changed much. On the other hand, the lower the similarity, the greater the difference between the historical path and the real-time path, and the greater the deviation of the real-time path from the historical path, indicating that the elderly person's path has changed significantly.

[0065] As an optional implementation, the step of identifying the elderly person's emotional state and assigning an emotional state label includes:

[0066] Collect video stream data captured by real-time monitoring cameras in the area where the elderly person is located, and divide the video stream data into frames to obtain image data;

[0067] The image data is processed using deep learning detection boxes to obtain the facial recognition results in each frame of the image, and the feature reference points of facial expressions are determined based on the facial recognition results.

[0068] A facial feature mesh is generated based on the feature reference points. Deformation is determined based on the facial feature mesh. The similarity between the deformation and the reference emotional feature mesh is calculated. The emotional state of the elderly person is determined based on the similarity calculation result, and an emotional state label is assigned.

[0069] Specifically, after locating the face using deep learning detection boxes, a facial feature grid is constructed based on the located face using feature reference points. The facial feature grid reflects the facial features. The model stores reference emotional feature grids for different emotional states. The facial feature grid is compared with the reference emotional feature grids, and the one with the highest similarity is taken as the elderly person's emotional state. For example, emotional states include joy, indifference, excitement, displeasure, sleep, and other states. After identifying the elderly person's emotional state, an emotional state label is assigned.

[0070] As an optional implementation, the health data includes heart rate, blood pressure, height, and weight;

[0071] Acquire real-time health data of the elderly, construct a health status recognition model, and input the basic health data, the real-time health data, the path similarity, and the emotional state label into the health status recognition model for training to obtain a trained health status recognition model, including:

[0072] The system acquires real-time health data of the elderly, calculates the rate of change of health data based on the real-time health data and the basic health data, and integrates the rate of change of health data with the real-time health data in chronological order to obtain a first health data feature set.

[0073] The first health data feature set is input into XGboot for feature filtering to obtain the second health data feature set;

[0074] A health status recognition model is constructed based on LSTM. The second health data feature set, the emotional state label, and the path similarity are input into the health status recognition model for training to obtain a trained health status recognition model.

[0075] Specifically, the rate of change of health data is calculated based on real-time and basic health data to identify abnormal fluctuations in health indicators, providing more meaningful input for subsequent model training. Health data and the rate of change are integrated into a first health data feature set, providing comprehensive input for subsequent feature selection and model training. XGBoost is used to select the most valuable features for health status identification, resulting in a second health data feature set, reducing the number of features and improving the model's training efficiency and generalization ability. LSTM, a neural network suitable for processing time-series data, can capture the dynamic changes of health data over time. The second health data feature set, emotional state labels, and path similarity are used to train the health status identification model, enabling it to accurately identify the health status of the elderly. The trained model is then used to identify real-time health data to obtain the identification results.

[0076] As an optional implementation, a second identifier is also included, which is bound to the mobile terminal of the administrator or the elderly person's family member, and the cloud platform stores the mapping relationship between the second identifier and the administrator's mobile terminal.

[0077] If abnormal behavior is detected, an early warning will be issued, including:

[0078] The cloud platform sends the location of the first identifier to the mobile terminal bound to the second identifier to issue an early warning.

[0079] Specifically, the second identifier is identical to the first identifier, both being active RFID tags. The second identifier is bound to the mobile terminal of the administrator or the elderly person's family member. When abnormal behavior is detected in the elderly person, the location detection unit obtains the specific location of the first identifier, i.e., the elderly person's real-time location. The cloud platform stores the mapping relationship between the first and second identifiers, thus enabling it to send this real-time location to the mobile terminal where the second identifier is located. In other words, the cloud platform sends the elderly person's real-time location to the administrator or family member. The cloud platform can communicate with the mobile terminal bound to the second identifier via WiFi or 4G / 5G, ensuring that when abnormal behavior is detected, the administrator or family member can be quickly notified, allowing them to quickly obtain the elderly person's real-time location and respond appropriately.

[0080] As an optional implementation, it also includes an electricity monitoring module, which is used to acquire the electricity consumption information of the indoor equipment when the elderly are in the indoor area, send the electricity consumption information to the cloud platform, the cloud platform analyzes the electricity consumption information, obtains the electricity consumption analysis results, determines whether the elderly's electricity consumption information is reasonable based on the electricity consumption analysis results, generates an electricity consumption message, and sends the electricity consumption message to the mobile terminal of the management personnel.

[0081] The process of acquiring indoor area device power consumption data, sending the power consumption data to a cloud platform, the cloud platform performing power consumption analysis on the data to obtain analysis results, determining whether the elderly person's power consumption information is reasonable based on the analysis results, generating a power consumption message, and sending the power consumption message to the administrator's mobile terminal includes:

[0082] Set a period for monitoring the elderly's electricity consumption, obtain electricity consumption information data during the monitoring period, and send the electricity consumption information data to the cloud platform;

[0083] The cloud platform stores preset electricity consumption thresholds for the elderly during the electricity monitoring period. The electricity consumption thresholds are compared with the electricity consumption information data. If the electricity consumption information data is greater than the electricity consumption thresholds, an electricity consumption growth rate analysis is performed.

[0084] If the electricity consumption data is less than the electricity consumption threshold, a normal electricity consumption message is generated and sent to the administrator's mobile terminal.

[0085] Specifically, the system acquires electricity consumption data for the indoor area where the elderly person resides during the electricity monitoring period. This indoor area can be the elderly person's rest area. During the day, the elderly person is more active and uses electrical appliances more frequently, so the electricity monitoring period can be set to daytime. The electricity consumption data during the monitoring period is sent to a cloud platform. The cloud platform stores and analyzes the electricity consumption data, comparing it with a preset electricity consumption threshold. The preset electricity consumption threshold is a reference value set based on factors such as the elderly person's living habits, appliance usage, and safe electricity use standards. For example, based on the frequency and power of the elderly person's daily appliance use and the general safe electricity use range, a reasonable upper limit for electricity consumption is set as the threshold. Exceeding this threshold may indicate abnormal electricity use. By comparing the electricity consumption data with the preset electricity consumption threshold, it is possible to quickly determine whether the elderly person's electricity use is within the normal range.

[0086] When electricity consumption data exceeds the electricity consumption threshold, it does not necessarily mean that there is an abnormality in electricity consumption. Therefore, further analysis of the electricity consumption growth rate is needed. The electricity consumption growth rate can reflect the trend of electricity consumption over a period of time. If the electricity consumption growth rate is abnormally high, it may mean that there are problems such as appliance failure, misoperation, or elderly people encountering emergencies (such as forgetting to turn off appliances).

[0087] When the electricity consumption data is less than the electricity consumption threshold, it indicates that the elderly person's electricity consumption is within the normal range. A normal electricity consumption message is generated and sent to the management personnel, providing normal feedback on the elderly person's electricity consumption and letting the management personnel know that the elderly person's electricity consumption is in a safe and reasonable state.

[0088] As an optional implementation, the electricity consumption growth rate analysis includes:

[0089] The power consumption data of each electrical device connected to the smart socket during the current power consumption monitoring period is obtained through the smart socket, and the power consumption data during the current power consumption monitoring period is sent to the cloud platform.

[0090] The cloud platform obtains the electricity consumption data from the previous electricity consumption monitoring period, calculates the difference between the electricity consumption data from the current electricity consumption monitoring period and the electricity consumption data from the previous electricity consumption monitoring period, and obtains the electricity consumption amplitude of each electrical device. The cloud platform presets an electricity consumption amplitude threshold, compares the electricity consumption amplitude with the electricity consumption amplitude threshold, and if the electricity consumption amplitude is greater than the electricity consumption amplitude threshold, generates an abnormal electricity consumption growth rate message and sends the abnormal electricity consumption growth rate message to the mobile terminal of the management personnel.

[0091] If the electricity consumption amplitude is less than the electricity consumption amplitude threshold, a normal electricity consumption growth rate message is generated and sent to the mobile terminal of the management personnel.

[0092] Specifically, smart sockets can accurately monitor the real-time power consumption of each connected device. Within a pre-defined monitoring period, acquiring power consumption data from these devices allows for accurate understanding of each device's power consumption during that time, providing foundational data for subsequent analysis of power consumption growth rates. The acquired power consumption data is then sent to a cloud platform, enabling centralized management and analysis of large amounts of data.

[0093] The system retrieves electricity consumption data from the previous monitoring period, which serves as a benchmark for the current period. Comparing this data with the current period's data clearly shows changes in electricity consumption, providing the necessary data foundation for calculating the electricity consumption growth rate. The cloud platform calculates the difference between the current and previous monitoring periods' electricity consumption data to obtain the electricity consumption amplitude for each device. This amplitude reflects the degree of change in each device's electricity consumption over different periods, and its magnitude indicates the extent of increase or decrease in electricity consumption. A preset electricity consumption amplitude threshold is a reference value set based on factors such as the elderly's normal electricity usage habits, appliance power, and safe electricity usage standards. It provides a clear standard for judging whether the electricity consumption amplitude is normal. By comparing the electricity consumption amplitude with the preset threshold, it can quickly determine whether the electricity consumption growth rate is within the normal range. When the electricity consumption amplitude exceeds the threshold, it indicates an abnormal electricity consumption growth rate, generating an abnormal electricity consumption growth rate message to promptly alert management personnel. Conversely, when the electricity consumption amplitude is less than the threshold, it indicates normal electricity usage, generating a normal electricity consumption growth rate message.

[0094] When anomalies are found by analyzing electricity consumption data, further analysis of the electricity consumption growth rate can be conducted to determine whether the electricity consumption data is abnormal. This can improve the accuracy of anomaly detection and avoid misjudgment.

[0095] Example 2

[0096] Please refer to the attached document. Figure 2 A cloud-based smart elderly care digital management method includes the following steps:

[0097] Obtain basic health data information of the elderly and send the basic health data information to the cloud platform for storage;

[0098] Obtain the elderly person's real-time location information, generate a real-time path based on the real-time location information, calculate the similarity between the real-time path and the historical path, and obtain the path similarity.

[0099] Facial recognition is performed on frame sequence images of surveillance video based on deep learning detection boxes. Based on the facial recognition results, the emotional state of the elderly is identified, and an emotional state label is assigned.

[0100] The elderly person's real-time health data is obtained, a health status recognition model is constructed, and the path similarity, emotional state label, and real-time health data are input into the health status recognition model for training to obtain the recognition result.

[0101] Based on the identification results, it is determined whether the elderly person exhibits abnormal behavior. If abnormal behavior is found, an early warning is issued.

[0102] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A cloud-based smart elderly care digital management system, characterized in that, include: The acquisition module is used to acquire the basic health data information of the elderly and send the basic health data information to the cloud platform; The path comparison module is used to obtain the elderly person's real-time location, generate the elderly person's real-time path based on the real-time location, and calculate the similarity between the real-time path and the historical path to obtain the path similarity. The emotion recognition module is used to identify the emotional state of the elderly and assign emotional state labels. The health status recognition module acquires the elderly's real-time health data information, constructs a health status recognition model, and inputs the basic health data information, the real-time health data information, the path similarity, and the emotional state label into the health status recognition model for training to obtain a trained health status recognition model. The elderly's health status is then recognized based on the trained health status recognition model to obtain the recognition result. The cloud platform determines whether the elderly person exhibits abnormal behavior based on the identification results, and issues an alert if abnormal behavior is detected.

2. The cloud-based smart elderly care digital management system according to claim 1, characterized in that, The path comparison module includes a first identifier, a location detection unit, and a path calculation unit; The first identifier stores the elderly person's basic information, and the first identifier is an active FRID tag; The location detection unit includes multiple units, which are evenly distributed in the activity area. The distance between each location detection unit and the first marker is calculated, and the distance is sent to the cloud platform. The cloud platform generates the elderly person's real-time path in chronological order. The path calculation unit is used to obtain the elderly person's historical path from the cloud platform, and calculate the similarity between the historical path and the real-time path based on Euclidean distance to obtain the path similarity.

3. The cloud-based smart elderly care digital management system according to claim 1, characterized in that, The process of identifying and labeling the emotional state of the elderly includes: Collect video stream data captured by real-time monitoring cameras in the area where the elderly person is located, and divide the video stream data into frames to obtain image data; The image data is processed using deep learning detection boxes to obtain the facial recognition results in each frame of the image, and the feature reference points of facial expressions are determined based on the facial recognition results. A facial feature mesh is generated based on the feature reference points. Deformation is determined based on the facial feature mesh. The similarity between the deformation and the reference emotional feature mesh is calculated. The emotional state of the elderly person is determined based on the similarity calculation result, and an emotional state label is assigned.

4. The cloud-based smart elderly care digital management system according to claim 1, characterized in that, The health data includes heart rate, blood pressure, height, and weight; Acquire real-time health data of the elderly, construct a health status recognition model, and input the basic health data, the real-time health data, the path similarity, and the emotional state label into the health status recognition model for training to obtain a trained health status recognition model, including: The system acquires real-time health data of the elderly, calculates the rate of change of health data based on the real-time health data and the basic health data, and integrates the rate of change of health data with the real-time health data in chronological order to obtain a first health data feature set. The first health data feature set is input into XGboot for feature filtering to obtain the second health data feature set; A health status recognition model is constructed based on LSTM. The second health data feature set, the emotional state label, and the path similarity are input into the health status recognition model for training to obtain a trained health status recognition model.

5. The cloud-based smart elderly care digital management system according to claim 2, characterized in that, It also includes a second identifier, which is bound to the mobile terminal of the administrator or the elderly person's family member. The cloud platform stores the mapping relationship between the second identifier and the administrator's mobile terminal. If abnormal behavior is detected, an early warning will be issued, including: The cloud platform sends the location of the first identifier to the mobile terminal bound to the second identifier to issue an early warning.

6. The cloud-based smart elderly care digital management system according to claim 1, characterized in that, It also includes an electricity monitoring module, which is used to acquire the electricity consumption information of the indoor equipment when the elderly are in the indoor area, send the electricity consumption information to the cloud platform, the cloud platform analyzes the electricity consumption information, obtains the electricity consumption analysis results, judges whether the elderly's electricity consumption information is reasonable based on the electricity consumption analysis results, generates an electricity consumption message, and sends the electricity consumption message to the mobile terminal of the management personnel.

7. The cloud-based smart elderly care digital management system according to claim 6, characterized in that, The process of acquiring indoor area device power consumption data, sending the power consumption data to a cloud platform, the cloud platform performing power consumption analysis on the data to obtain analysis results, determining whether the elderly person's power consumption information is reasonable based on the analysis results, generating a power consumption message, and sending the power consumption message to the administrator's mobile terminal includes: Set a period for monitoring the elderly's electricity consumption, obtain electricity consumption information data during the monitoring period, and send the electricity consumption information data to the cloud platform; The cloud platform stores preset electricity consumption thresholds for the elderly during the electricity monitoring period. The electricity consumption thresholds are compared with the electricity consumption information data. If the electricity consumption information data is greater than the electricity consumption thresholds, an electricity consumption growth rate analysis is performed. If the electricity consumption data is less than the electricity consumption threshold, a normal electricity consumption message is generated and sent to the administrator's mobile terminal.

8. The cloud-based smart elderly care digital management system according to claim 7, characterized in that, The electricity consumption growth rate analysis includes: The power consumption data of each electrical device connected to the smart socket during the current power consumption monitoring period is obtained through the smart socket, and the power consumption data during the current power consumption monitoring period is sent to the cloud platform. The cloud platform obtains the electricity consumption data from the previous electricity consumption monitoring period, calculates the difference between the electricity consumption data from the current electricity consumption monitoring period and the electricity consumption data from the previous electricity consumption monitoring period, and obtains the electricity consumption amplitude of each electrical device. The cloud platform presets an electricity consumption amplitude threshold, compares the electricity consumption amplitude with the electricity consumption amplitude threshold, and if the electricity consumption amplitude is greater than the electricity consumption amplitude threshold, generates an abnormal electricity consumption growth rate message and sends the abnormal electricity consumption growth rate message to the mobile terminal of the management personnel. If the electricity consumption amplitude is less than the electricity consumption amplitude threshold, a normal electricity consumption growth rate message is generated and sent to the mobile terminal of the management personnel.

9. A cloud-based smart elderly care digital management method, characterized in that, Includes the following steps: Obtain basic health data information of the elderly and send the basic health data information to the cloud platform for storage; Obtain the elderly person's real-time location information, generate a real-time path based on the real-time location information, calculate the similarity between the real-time path and the historical path, and obtain the path similarity. Facial recognition is performed on frame sequence images of surveillance video based on deep learning detection boxes. Based on the facial recognition results, the emotional state of the elderly is identified, and an emotional state label is assigned. The elderly person's real-time health data is obtained, a health status recognition model is constructed, and the path similarity, emotional state label, and real-time health data are input into the health status recognition model for training to obtain the recognition result. Based on the identification results, it is determined whether the elderly person exhibits abnormal behavior. If abnormal behavior is found, an early warning is issued.

Images