An internet-of-things-based intelligent pension service management system

Abstract

The present application relates to the field of old-age service management, and particularly relates to a smart old-age service management system based on the Internet of Things, which is provided with an observation and collection module, which comprises an image collection unit for collecting infrared images of a monitoring target and an audio collection unit for collecting respiratory sound volumes of the monitoring target; a behavior analysis module, which is used to identify respiratory behavior characteristics based on the infrared images of the monitoring target every monitoring period, analyze respiratory state abnormality characteristic values for the monitoring target based on the respiratory behavior characteristics, and divide respiratory state abnormality tendency categories of the monitoring target; and a behavior analysis module, which is used to adaptively analyze the monitoring target according to the respiratory state abnormality tendency categories. The present application analyzes the characteristic state presented by the monitoring target during night sleep, discovers the sleep abnormality of the monitoring target in time, issues a warning prompt information so as to enable a caretaker to check in time, ensures the sleep health of the monitoring target, and reduces the risk of occurrence of other health problems.

Description

Technical Field

[0001] This invention relates to the field of elderly care service management, and in particular to a smart elderly care service management system based on the Internet of Things. Background Technology

[0002] With the changing structure of the elderly population, the demand for elderly care services at different levels and with different needs is becoming increasingly diversified. As an important component of the new elderly care service model, the elderly care service management system improves the efficiency and quality of elderly care services through digital and intelligent means, meeting the diversified needs of the market. In particular, the sleep quality of the elderly gradually decreases with age, and there may even be some abnormal risks in sleep conditions. For example, the elderly are prone to breathing difficulties at night, which requires caregivers to check them in a timely manner.

[0003] Chinese Patent Application Publication No. CN117877750A discloses an IoT-based smart elderly care service management system. This invention obtains the local correlation between each elderly person and each pair of body indicator curves; then obtains the overall correlation between each pair of body indicators; based on the overall correlation between body indicators, it selects highly correlated indicator pairs to obtain the correlation between each body indicator and other body indicators; combining the number of corresponding highly correlated indicator pairs, it obtains the weight value of each body indicator; at each time point, based on the differences in body indicator data between each elderly person in each dimension and the corresponding weight value of the body indicator, it obtains the distance between each elderly person's body indicators; and it clusters all elderly people for elderly care service management.

[0004] However, the following problems still exist in the existing technology.

[0005] As people age, their cardiopulmonary function and sleep structure change. In the process of caring for the elderly, it is easy to overlook subtle changes in their breathing and sleep quality during nighttime sleep. At the same time, caregivers cannot watch over the elderly at all times. Therefore, abnormal conditions such as difficulty breathing may occur during nighttime sleep. For example, excessively large breathing amplitude and excessive number and changes in limb movements may occur during nighttime sleep. It is impossible to ensure that timely examination and adjustment can be made when abnormal conditions occur in the elderly. Summary of the Invention

[0006] To address this, the present invention provides an IoT-based smart elderly care service management system to overcome the problems in existing technologies where, as the elderly age, their cardiopulmonary function and sleep structure change. During the care of the elderly, subtle changes in their breathing and sleep quality during nighttime sleep are easily overlooked. Furthermore, caregivers cannot constantly monitor the elderly, leading to potential breathing difficulties and other abnormalities during sleep, and timely checks and adjustments cannot be guaranteed when these abnormalities occur.

[0007] To achieve the above objectives, the present invention provides a smart elderly care service management system based on the Internet of Things, comprising:

[0008] The observation and acquisition module includes an image acquisition unit for acquiring infrared images of the monitoring target and an audio acquisition unit for acquiring the breathing volume of the monitoring target;

[0009] The behavior analysis module, which is connected to the observation and acquisition module, is used to identify respiratory behavior characteristics based on the infrared image of the monitored target at each monitoring cycle, including the number of limb movements and the respiratory amplitude. Based on the respiratory behavior characteristics, it analyzes the abnormal respiratory state characterization value of the monitored target to classify the abnormal respiratory state tendency category of the monitored target.

[0010] A behavior analysis module, connected to both the observation and acquisition module and the behavior analysis module, is used to analyze the monitored target based on the category of abnormal respiratory state tendencies, including:

[0011] The breathing amplitude determined by the behavior analysis module is continuously invoked to construct a breathing amplitude time-domain curve for the monitored target. Abnormal breathing time-domain segments are identified based on the abrupt change characteristics of the breathing amplitude time-domain curve. The breathing volume of the monitored target within the abnormal breathing time-domain segment is combined with the frequency of limb movements to determine the state risk characterization coefficient, so as to determine whether to issue a warning message.

[0012] Alternatively, determine the number of sleep episodes and sleep duration of the monitored target at night, and adjust the monitoring cycle of the behavior analysis module.

[0013] Furthermore, the process by which the behavior analysis module identifies respiratory behavior characteristics based on the infrared image of the monitored target includes:

[0014] Used to identify and monitor target contours based on infrared images;

[0015] Used to identify whether limb movements occur based on the dynamic changes of the monitored target contour within a reference time period, and to determine the number of limb movements;

[0016] The maximum gas boundary contour of a single exhaled gas from a monitoring target is identified based on the infrared image, and the area of ​​the maximum gas boundary contour is determined as the breathing amplitude.

[0017] Furthermore, the process by which the behavior analysis module analyzes abnormal respiratory state representation values ​​for the monitored target based on the respiratory behavior characteristics includes:

[0018] The ratio of the number of limb movements of the monitored target to the threshold number of limb movements is used as the first respiratory state feature.

[0019] The ratio of the respiratory amplitude of the monitored target to the respiratory amplitude threshold is used as a second respiratory state feature to determine the respiratory state feature.

[0020] The first respiratory state feature and the second respiratory state feature are weighted and summed to determine the respiratory state abnormality characterization value.

[0021] Furthermore, the behavior analysis module is used to classify the abnormal respiratory state tendencies of the monitored target into categories, including:

[0022] If the abnormal respiratory state characterization value is greater than or equal to the abnormal respiratory state characterization threshold, the abnormal respiratory state tendency of the monitored target is determined to be a high tendency category.

[0023] If the abnormal respiratory state characterization value is less than the abnormal respiratory state characterization threshold, the abnormal respiratory state tendency of the monitored target is determined to be low tendency category.

[0024] Furthermore, the behavior analysis module is used to analyze the monitoring target based on the abnormal respiratory state tendency category, including,

[0025] If the respiratory state of the monitored target is in the high-probability category, the respiratory amplitude determined by the behavior analysis module is continuously invoked to construct a respiratory amplitude time-domain curve for the monitored target. The abnormal respiratory time-domain segment is identified based on the abrupt change characteristics of the respiratory amplitude time-domain curve. The respiratory volume of the monitored target in the abnormal respiratory time-domain segment is obtained and the frequency of limb movements is combined to determine the state risk characterization coefficient, so as to determine whether to issue a warning message.

[0026] If the respiratory status of the monitored target shows a low tendency for abnormality, then the number of sleeps and the duration of sleep at night are determined, and the monitoring cycle of the behavior analysis module is adjusted.

[0027] Furthermore, the process by which the behavior analysis module constructs a time-domain curve of respiratory amplitude for the monitored target includes:

[0028] This is used to construct a rectangular coordinate system with time as the horizontal axis and breathing amplitude as the vertical axis;

[0029] Coordinate points used to mark the respiratory amplitude at each moment in the rectangular coordinate system;

[0030] This is used to connect the coordinate points using a smooth curve to obtain the time-domain curve of the respiratory amplitude.

[0031] Furthermore, the process by which the behavior analysis module identifies abnormal breathing time-domain segments based on the abrupt change characteristics of the breathing amplitude time-domain curve includes:

[0032] This is used to compare the mean respiratory amplitude of each sub-time domain segment corresponding to the respiratory amplitude time domain curve with the mean respiratory amplitude of the corresponding segment in the adjacent time domain segment, and to solve for the respiratory amplitude difference.

[0033] If the respiratory amplitude difference corresponding to any sub-time domain segment is greater than or equal to the respiratory amplitude mutation threshold, then the sub-time domain segment is determined to have mutation characteristics and is identified as a respiratory abnormality time domain segment.

[0034] Furthermore, the process by which the behavior analysis module determines the state risk representation coefficients includes,

[0035] The ratio of the respiratory volume of the monitored target to the respiratory volume threshold is used as the first risk characteristic;

[0036] The ratio of the frequency of limb movements of the monitored target during the respiratory abnormality time domain to the limb movement frequency threshold is used as a second risk feature.

[0037] The sum of the first risk feature and the second risk feature is used to determine the state risk characterization coefficient.

[0038] Furthermore, the behavior analysis module is used to determine whether to issue a warning message, including,

[0039] If the state risk characterization coefficient is greater than or equal to the state risk characterization coefficient threshold, an early warning message will be issued.

[0040] Furthermore, the behavior analysis module is used to adjust the monitoring cycle of the behavior analysis module, including:

[0041] Increase the monitoring cycle of the behavior analysis module.

[0042] Compared with existing technologies, this invention sets up an observation and acquisition module, which includes an image acquisition unit for acquiring infrared images of the monitored target and an audio acquisition unit for acquiring the respiratory volume of the monitored target; a behavior analysis module, which is connected to the observation and acquisition module, is used to identify respiratory behavior characteristics based on the infrared images of the monitored target at each monitoring cycle, and analyze the abnormal respiratory state characterization values ​​of the monitored target based on the respiratory behavior characteristics to classify the abnormal respiratory state tendency category of the monitored target; a behavior analysis module, which is connected to both the observation and acquisition module and the behavior analysis module, is used to adaptively analyze the monitored target according to the abnormal respiratory state tendency category. This invention analyzes the characteristic state of the monitored target during nighttime sleep, so as to promptly detect abnormal sleep conditions of the monitored target, issue early warning information so that caregivers can check in a timely manner, ensure the sleep health of the monitored target, and reduce the risk of other health problems.

[0043] In particular, this invention analyzes the abnormal respiratory state characteristics of the monitored target based on respiratory behavior features. In reality, as the monitored target ages, their cardiopulmonary function may gradually decline, which may lead to breathing difficulties during sleep. This is usually manifested as a longer duration of a single breath compared to a normal breath, and more air being exhaled during a single breath. Furthermore, the monitored target may also exhibit increased limb movements during sleep at night, thereby affecting the monitored target's sleep quality and potentially leading to other health problems. Therefore, this application uses abnormal respiratory state characteristics to characterize the degree of abnormal respiratory state tendencies exhibited by the monitored target during sleep at night, providing data support for subsequent classification of abnormal respiratory state tendencies of the monitored target, and thus adaptively analyzing the monitored target.

[0044] In particular, this invention provides targeted analysis for monitoring targets with different categories of abnormal breathing tendencies. For monitoring targets with a high tendency for abnormal breathing tendencies, a time-domain curve of respiratory amplitude is constructed based on the target's breathing amplitude during nighttime sleep for analysis. During the actual sleep process, the target may experience a sudden and abnormal increase in respiratory amplitude due to breathing difficulties. In this case, the target's breathing will be accompanied by relatively loud snoring. At the same time, the target will make more limb movements when snoring frequently, breathing is not smooth, and the target is unable to adjust to a comfortable sleep state. The state risk characterization coefficient of the monitoring target is determined based on the characteristics presented in the above situations to characterize the degree of discomfort and abnormality of the target's current sleep and the degree of risk of sleep problems, so as to issue timely warning information so that caregivers can check the monitoring target in a timely manner; for those with a high tendency for abnormal breathing tendencies... For monitoring targets with a low tendency to exhibit abnormal breathing patterns during nighttime sleep, their sleep quality is generally good due to their consistently good breathing patterns over a long historical observation period. This allows for adaptive adjustments to the monitoring cycle, saving computational resources. Furthermore, as monitoring targets age, their sleep patterns change, primarily manifested as multiple brief awakenings during nighttime sleep and a shortening of total sleep duration. Over time, this can negatively impact sleep quality. Therefore, this invention obtains the number of sleep episodes and sleep duration of monitoring targets. It specifically analyzes the characteristic states exhibited by monitoring targets with different breathing pattern abnormalities during nighttime sleep. By analyzing the target's sleep and breathing states, it identifies sleep abnormalities and issues early warning messages to facilitate timely checks by caregivers, ensuring the monitoring targets' sleep health and reducing the risk of other health problems. Attached Figure Description

[0045] Figure 1 A functional block diagram of an IoT-based smart elderly care service management system, as an embodiment of the invention;

[0046] Figure 2 A logic diagram for classifying abnormal respiratory state tendencies of monitoring targets in an embodiment of the invention;

[0047] Figure 3 A logic decision diagram for identifying time-domain segments of respiratory abnormalities in an embodiment of the invention;

[0048] Figure 4 This is a logic diagram for determining whether to issue a warning message in an embodiment of the invention. Detailed Implementation

[0049] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0050] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0051] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral connection; it can refer to a mechanical connection or an electrical connection. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0052] Please see Figures 1 to 4 As shown, Figure 1 This is a functional block diagram of the IoT-based smart elderly care service management system according to an embodiment of the present invention. Figure 2 This is a logic diagram for classifying abnormal respiratory state tendencies of monitored targets according to embodiments of the present invention. Figure 3 This is a logic decision diagram for identifying abnormal breathing time domain segments in an embodiment of the invention. Figure 4 This is a logic diagram illustrating whether to issue a warning message in an embodiment of the invention. The IoT-based smart elderly care service management system of this invention includes:

[0053] The observation and acquisition module includes an image acquisition unit for acquiring infrared images of the monitoring target and an audio acquisition unit for acquiring the breathing volume of the monitoring target;

[0054] The behavior analysis module, which is connected to the observation and acquisition module, is used to identify respiratory behavior characteristics based on the infrared image of the monitored target at each monitoring cycle, including the number of limb movements and the respiratory amplitude. Based on the respiratory behavior characteristics, it analyzes the abnormal respiratory state characterization value of the monitored target to classify the abnormal respiratory state tendency category of the monitored target.

[0055] A behavior analysis module, connected to both the observation and acquisition module and the behavior analysis module, is used to analyze the monitored target based on the category of abnormal respiratory state tendencies, including:

[0056] The breathing amplitude determined by the behavior analysis module is continuously invoked to construct a breathing amplitude time-domain curve for the monitored target. Abnormal breathing time-domain segments are identified based on the abrupt change characteristics of the breathing amplitude time-domain curve. The breathing volume of the monitored target within the abnormal breathing time-domain segment is combined with the frequency of limb movements to determine the state risk characterization coefficient, so as to determine whether to issue a warning message.

[0057] Alternatively, determine the number of sleep episodes and sleep duration of the monitored target at night, and adjust the monitoring cycle of the behavior analysis module.

[0058] Understandably, the monitoring target is the human body, and in practice, the monitoring target is the elderly in nursing homes.

[0059] Specifically, there are no restrictions on the specific structure of the image acquisition unit, as long as it has the function of acquiring infrared images of the monitored target. For example, high-definition infrared cameras can be installed in the rooms of each monitored target to acquire images. Of course, other methods can also be used, which will not be elaborated here.

[0060] Specifically, there are no restrictions on the specific structure of the audio acquisition unit, as long as it has the function of acquiring the breathing volume of the target. For example, it can be a sound receiving device. In practice, the sound receiving device is placed in the sleeping area of ​​the target, such as on the bedside table, to acquire audio. Of course, other methods can also be used, which will not be elaborated here.

[0061] Specifically, there are no limitations on the behavior analysis module and the behavior parsing module or their individual units. They can be composed of logic components or combinations of logic components, including field-programmable processors, computers or microprocessors in computers.

[0062] Specifically, the process by which the behavior analysis module identifies respiratory behavior characteristics based on the infrared image of the monitored target includes:

[0063] Used to identify and monitor target contours based on infrared images;

[0064] Used to identify whether limb movements have occurred based on the dynamic changes of the monitored target contour within a reference time period, and to determine the number of limb movements. The reference time period should be less than the monitoring cycle.

[0065] The maximum gas boundary contour of a single exhaled gas from a monitoring target is identified based on the infrared image, and the area of ​​the maximum gas boundary contour is determined as the breathing amplitude.

[0066] In this embodiment, there is no limitation on the method of identifying dynamic changes in the outline of the monitored target. For example, by setting several marker points on the outline of the infrared image of the monitored target, the average displacement of each marker point is determined. If the average displacement is greater than the displacement threshold, it is determined that a limb movement has occurred. In order to reflect the amount of movement, the displacement threshold must be greater than 2cm.

[0067] It is understandable that since human respiration generates heat, the exhaled gas can be identified by the infrared acquisition unit and thus reflected in the infrared image. For example, the boundary contour of a single exhaled gas from the monitoring target can be determined by edge recognition algorithm or image clustering algorithm. Of course, other methods can also be used, which will not be elaborated here.

[0068] Specifically, the process by which the behavior analysis module analyzes abnormal respiratory state representation values ​​for the monitored target based on the respiratory behavior characteristics includes:

[0069] The ratio of the number of limb movements of the monitored target to the threshold number of limb movements is used as the first respiratory state feature.

[0070] The ratio of the respiratory amplitude of the monitored target to the respiratory amplitude threshold is used as a second respiratory state feature to determine the respiratory state feature.

[0071] The first respiratory state feature and the second respiratory state feature are weighted and summed to determine the respiratory state abnormality characterization value.

[0072] In this implementation, when performing weighted summation, the weight of the first respiratory state feature is set to 0.45, and the weight of the second respiratory state feature is set to 0.55.

[0073] The thresholds for the number of limb movements and the respiratory amplitude of the monitored targets are preset. The respiratory behavior characteristics of several monitored targets during sleep are obtained. The data on the number of limb movements and the respiratory amplitude are retrieved, and the mean of the number of limb movements and the mean of the respiratory amplitude are calculated. The threshold for the number of limb movements is set to be 1.01 to 1.04 times the mean of the number of limb movements, and the threshold for the respiratory amplitude is set to be 1.05 to 1.08 times the mean of the respiratory amplitude.

[0074] This invention analyzes abnormal respiratory state characteristics of monitoring targets based on respiratory behavior features. In reality, as people age, the cardiopulmonary function of monitoring targets may gradually decline, which may lead to breathing difficulties during sleep. This is usually manifested as a longer duration of a single breath compared to normal breathing, and more air being exhaled during a single breath. In addition, the monitoring targets may also exhibit more limb movements during sleep at night, which may affect the quality of sleep and potentially lead to other health problems. Therefore, this application uses abnormal respiratory state characteristics to characterize the degree of abnormal respiratory state tendencies exhibited by monitoring targets during sleep at night, providing data support for subsequent classification of abnormal respiratory state tendencies of monitoring targets, and thus enabling adaptive analysis of monitoring targets.

[0075] Specifically, the behavior analysis module is used to classify the abnormal respiratory state tendencies of the monitored target, including,

[0076] If the abnormal respiratory state characterization value is greater than or equal to the abnormal respiratory state characterization threshold, the abnormal respiratory state tendency of the monitored target is determined to be a high tendency category.

[0077] If the abnormal respiratory state characterization value is less than the abnormal respiratory state characterization threshold, the abnormal respiratory state tendency of the monitored target is determined to be low tendency category.

[0078] The threshold for abnormal respiratory status is selected within the range [1.23, 1.37].

[0079] Specifically, the behavior analysis module is used to analyze the monitored target based on the category of abnormal respiratory state tendencies, including,

[0080] If the respiratory state of the monitored target is in the high-probability category, the respiratory amplitude determined by the behavior analysis module is continuously invoked to construct a respiratory amplitude time-domain curve for the monitored target. The abnormal respiratory time-domain segment is identified based on the abrupt change characteristics of the respiratory amplitude time-domain curve. The respiratory volume of the monitored target in the abnormal respiratory time-domain segment is obtained and the frequency of limb movements is combined to determine the state risk characterization coefficient, so as to determine whether to issue a warning message.

[0081] If the respiratory status of the monitored target shows a low tendency for abnormality, then the number of sleeps and the duration of sleep at night are determined, and the monitoring cycle of the behavior analysis module is adjusted.

[0082] Specifically, the process by which the behavior analysis module constructs a time-domain curve of respiratory amplitude for the monitored target includes:

[0083] This is used to construct a rectangular coordinate system with time as the horizontal axis and breathing amplitude as the vertical axis;

[0084] Coordinate points used to mark the respiratory amplitude at each moment in the rectangular coordinate system;

[0085] This is used to connect the coordinate points using a smooth curve to obtain the time-domain curve of the respiratory amplitude.

[0086] Specifically, there are no restrictions on the method for constructing the respiratory amplitude time-domain curve. For example, the time-domain curve can be fitted using MATLAB correlation fitting software, which will not be elaborated further.

[0087] Specifically, the process by which the behavior analysis module identifies abnormal breathing time-domain segments based on the abrupt change characteristics of the breathing amplitude time-domain curve includes:

[0088] This is used to compare the mean respiratory amplitude of each sub-time domain segment corresponding to the respiratory amplitude time domain curve with the mean respiratory amplitude of the corresponding segment in the adjacent time domain segment, and solve for the respiratory amplitude difference. It can be understood that there can be two adjacent time domain segments. First, the mean of the respiratory amplitude in the adjacent time domain segment is solved.

[0089] If the respiratory amplitude difference corresponding to any sub-time domain segment is greater than or equal to the respiratory amplitude mutation threshold, then the sub-time domain segment is determined to have mutation characteristics and is identified as a respiratory abnormality time domain segment.

[0090] The threshold for sudden changes in respiratory amplitude is 0.3 times the threshold for respiratory amplitude.

[0091] Specifically, the process by which the behavior analysis module determines the state risk representation coefficients includes,

[0092] The ratio of the respiratory volume of the monitored target to the respiratory volume threshold is used as the first risk characteristic;

[0093] The ratio of the frequency of limb movements of the monitored target during the respiratory abnormality time domain to the limb movement frequency threshold is used as a second risk feature.

[0094] The sum of the first risk feature and the second risk feature is used to determine the state risk characterization coefficient.

[0095] In this embodiment, the respiratory volume threshold of the monitored target and the frequency of limb movements of the monitored target during the abnormal respiratory time domain are preset. The respiratory behavior characteristics and respiratory volume of several monitored targets during sleep are obtained. The respiratory volume data and the frequency of limb movements of the monitored target during the abnormal respiratory time domain are called up to solve for the average respiratory volume and the average frequency of limb movements. The respiratory volume threshold is set to be 1.12 to 1.22 times the average respiratory volume, and the limb movement frequency threshold is set to be 1.04 to 1.18 times the average limb movement frequency.

[0096] Specifically, the behavior analysis module is used to determine whether to issue a warning message, including,

[0097] If the state risk characterization coefficient is greater than or equal to the state risk characterization coefficient threshold, an early warning message will be issued.

[0098] If the state risk characterization coefficient is less than the state risk characterization coefficient threshold, there is no need to issue an early warning message.

[0099] The threshold for the state risk characterization coefficient is selected within the interval [2.23, 2.42].

[0100] In this embodiment, the warning information includes the name of the monitored target, the room where the monitored target is located, the current respiratory behavior characteristics of the monitored target, the breathing volume of the monitored target, and the frequency of limb movements of the monitored target when it is in the abnormal breathing time domain. The warning information can be sent to the monitoring terminal.

[0101] Specifically, the behavior analysis module is used to adjust the monitoring cycle of the behavior analysis module, including:

[0102] Increase the monitoring cycle of the behavior analysis module.

[0103] It is understandable that the monitoring cycle of the observation and acquisition module can be adjusted because when the tendency for abnormal respiratory states is low, the sleep state is stable. Therefore, the monitoring cycle can be appropriately extended to save computing power. The delayed monitoring cycle can be 1.5 to 2 times the initial monitoring cycle, which can be selected within the range [10s, 30s].

[0104] In this embodiment, for monitoring targets with a low tendency to abnormal breathing status, a sleep quality report is generated in a timely manner. The sleep quality includes the number of nighttime sleeps and the duration of nighttime sleep for the monitoring target.

[0105] Specifically, the identification of sleep state can be determined based on the respiratory behavior characteristics of the monitored target. For example, the historical average respiratory amplitude and the historical number of limb movements during the monitored target's sleep state can be obtained. The difference ratio of the average respiratory amplitude obtained in real time and the historical average respiratory amplitude can be calculated. The difference ratio of the number of limb movements obtained in real time and the historical number of limb movements can be calculated. If the average difference ratio is within 0.1 to 0.15, the target is determined to be in a sleep state. If the difference ratio of the average respiratory amplitude and / or the difference ratio of the number of movements are not within 0.1 to 0.15, the target is determined not to be in a sleep state. The number of sleep episodes and the duration of sleep can be identified based on this. This will not be elaborated further.

[0106] This invention provides targeted analysis for monitoring targets with different categories of abnormal respiratory states. For monitoring targets with a high tendency to exhibit abnormal respiratory states, a time-domain curve of respiratory amplitude is constructed based on the target's respiratory amplitude during nighttime sleep for analysis. During actual sleep, the target may experience a sudden and abnormal increase in respiratory amplitude due to breathing difficulties. In this case, the target's breathing will be accompanied by relatively loud snoring. Simultaneously, the target will exhibit more limb movements when snoring frequently, breathing is difficult, and the target is unable to adjust to a comfortable sleep state. Based on the characteristics presented in the above situations, a state risk characterization coefficient is determined to represent the current degree of sleep discomfort and abnormality, as well as the risk of sleep problems, so as to issue timely warning information for caregivers to check on the target promptly. For monitoring targets with a high tendency to exhibit abnormal respiratory states... For monitoring targets with low propensity for abnormal breathing, their sleep quality is generally good due to their consistently good breathing patterns during nighttime sleep over a long period of historical observation. This allows for adaptive adjustments to the monitoring cycle, saving computational resources. Furthermore, as monitoring targets age, their sleep patterns change, primarily manifested as multiple brief awakenings during nighttime sleep and a shortening of total sleep duration. Over time, this can negatively impact sleep quality. Therefore, this invention obtains the number of sleep episodes and sleep duration of monitoring targets. It specifically analyzes the characteristic states of monitoring targets with different respiratory propensity for abnormal breathing patterns during nighttime sleep. By analyzing the target's sleep and breathing states, it identifies sleep abnormalities and issues early warning messages to facilitate timely checks by caregivers, ensuring the monitoring targets' sleep health and reducing the risk of other health problems.

[0107] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An Internet of Things-based smart elderly care service management system, characterized in that, The method comprises the following steps: an observation collection module comprising an image collection unit for collecting infrared images of a monitoring target and an audio collection unit for collecting respiratory sound volume of the monitoring target; a behavior analysis module connected with the observation collection module for identifying respiratory behavior features including limb movement frequency and respiratory amplitude based on the infrared images of the monitoring target every monitoring period, analyzing respiratory state abnormality characteristic values of the monitoring target based on the respiratory behavior features, and classifying respiratory state abnormality tendency categories of the monitoring target; a behavior analysis module connected with the observation collection module and the behavior analysis module for analyzing the monitoring target according to the respiratory state abnormality tendency categories, including, continuously calling the respiratory amplitude determined by the behavior analysis module to construct a respiratory amplitude time domain curve of the monitoring target, identifying a respiratory abnormality time domain segment according to mutation characteristics of the respiratory amplitude time domain curve, obtaining state risk characteristic coefficients of the monitoring target in the respiratory abnormality time domain segment by combining respiratory sound volume and limb movement frequency to determine whether to send a warning prompt information; or, determining the number of sleep times and sleep duration of the monitoring target at night, and adjusting the monitoring period of the behavior analysis module; the process of the behavior analysis module for identifying respiratory behavior features based on the infrared images of the monitoring target comprises, identifying the outline of the monitoring target based on the infrared images; identifying whether limb movement occurs according to the dynamic changes of the outline of the monitoring target within a reference time period, and determining the frequency of limb movement; identifying the maximum gas boundary outline of the exhaled gas of the monitoring target according to the infrared images, and determining the area of the maximum gas boundary outline as the respiratory amplitude; the process of the behavior analysis module for determining state risk characteristic coefficients comprises, determining the ratio of respiratory sound volume of the monitoring target to respiratory sound volume threshold as a first risk feature; determining the ratio of the frequency of limb movement of the monitoring target in the respiratory abnormality time domain segment to the frequency of limb movement threshold as a second risk feature; determining the sum of the first risk feature and the second risk feature as the state risk characteristic coefficient. 2.The Internet of Things-based smart aged care service management system according to claim 1, characterized in that, the process of the behavior analysis module for analyzing respiratory state abnormality characteristic values of the monitoring target based on the respiratory behavior features comprises, determining the ratio of the frequency of limb movement of the monitoring target to the frequency of limb movement threshold as a first respiratory state feature; determining the ratio of the respiratory amplitude of the monitoring target to the respiratory amplitude threshold as a second respiratory state feature; determining the weighted sum of the first respiratory state feature and the second respiratory state feature as the respiratory state abnormality characteristic value. 3.The Internet of Things-based smart pension service management system according to claim 1, characterized in that, the process of the behavior analysis module for classifying respiratory state abnormality tendency categories of the monitoring target comprises, if the respiratory state abnormality characteristic value is greater than or equal to the respiratory state abnormality characteristic threshold, it is determined that the respiratory state abnormality tendency of the monitoring target is a high tendency category; if the respiratory state abnormality characteristic value is less than the respiratory state abnormality characteristic threshold, it is determined that the respiratory state abnormality tendency of the monitoring target is a low tendency category. 4.The Internet of Things based smart service management system for the elderly according to claim 1, wherein, The behavior analysis module is used to analyze the monitoring target according to the respiratory state abnormal tendency category, comprises, if the respiratory state abnormal tendency of the monitoring target is the high tendency category, the respiratory amplitude determined by the behavior analysis module is continuously called to construct the respiratory amplitude time domain curve of the monitoring target, the respiratory abnormal time domain segment is identified according to the mutation characteristics of the respiratory amplitude time domain curve, the state risk representation coefficient is obtained by combining the respiratory volume of the monitoring target in the respiratory abnormal time domain segment with the limb action occurrence frequency to determine whether the early warning prompt information is sent; if the respiratory state abnormal tendency of the monitoring target is the low tendency category, the sleep times and sleep duration of the monitoring target at night are determined, and the monitoring period of the behavior analysis module is adjusted. 5.The Internet-of-Things based smart service management system for the elderly according to claim 1, wherein, The process of constructing the respiratory amplitude time domain curve of the monitoring target by the behavior analysis module comprises, a rectangular coordinate system is constructed with time as the horizontal axis and respiratory amplitude as the vertical axis; the coordinate points of respiratory amplitude at each time are calibrated in the rectangular coordinate system; the respiratory amplitude time domain curve is obtained by connecting each coordinate point through a smooth curve. 6.The Internet-of-Things based smart service management system for the elderly according to claim 1, wherein, The process of identifying the respiratory abnormal time domain segment according to the mutation characteristics of the respiratory amplitude time domain curve by the behavior analysis module comprises, the respiratory amplitude difference value is solved by comparing the respiratory amplitude mean value of each sub time domain segment of the respiratory amplitude time domain curve with the corresponding respiratory amplitude mean value in the adjacent time domain segment; if the respiratory amplitude difference value of any sub time domain segment is greater than or equal to the respiratory amplitude mutation threshold value, it is determined that the sub time domain segment has mutation characteristics, and the sub time domain segment is determined as the respiratory abnormal time domain segment. 7.The Internet-of-Things based smart service management system for the elderly according to claim 1, wherein, The process of determining whether the early warning prompt information is sent by the behavior analysis module comprises, if the state risk representation coefficient is greater than or equal to the state risk representation coefficient threshold value, the early warning prompt information is sent. 8.The Internet-of-Things based smart service management system for the elderly according to claim 1, wherein, The process of adjusting the monitoring period of the behavior analysis module by the behavior analysis module comprises, the monitoring period of the behavior analysis module is increased.

Images