Remote home health state multi-dimensional monitoring system and method

Abstract

The invention provides a remote home health state multi-dimensional monitoring system and method, and belongs to the technical field of remote communication and health monitoring. The system comprises a data acquisition layer, a data control layer, a data analysis layer and a data output layer. The data control layer comprises a first data queue corresponding to the image sensor and a second data queue corresponding to the sound sensor; the data analysis layer analyzes at least part of sensing data stored in the first data queue and / or the second data queue; and the data output layer receives the analysis result of the data analysis layer, outputs the remote home health state of the target person, and sends a feedback indication to the wearable device worn by the target person when the remote home health state of the target person meets a preset standard. The method is implemented based on the system. According to the method, remote home health state comprehensive judgment and monitoring can be carried out based on multi-dimensional data, different dimensions are mutually complemented and activated, and the data energy consumption of the sensor can be reduced while the accuracy can be ensured.

Description

Technical Field

[0001] The present invention belongs to the technical field of remote communication and health monitoring, and in particular, relates to a remote home health status multi-dimensional monitoring system and method, a computer-readable storage medium for implementing the method, a computer program product and an electronic device. Background Art

[0002] With the acceleration of the aging of the population in my country, remote home health services have become a hot topic. The traditional home service model can no longer meet the life needs of home users. By applying communication technology, information technology, Internet of Things technology, artificial intelligence, etc. to the home health service environment, an intelligent and personalized service system is built to provide home users with a comfortable, convenient, safe and controllable home health service environment, which has become a new direction for the development of remote home health.

[0003] Remote home health care services integrate medical equipment with communication technology and apply them to health, medical and care services to reduce the need for human care. Chinese invention patent CN113034850B provides an intelligent home health management device that can control and interact with health-related factors and the surrounding environment, so that patients are always in a good health management state at home; Chinese invention patent CN113113145B discloses an integrated intelligent platform for smart home elderly care service management based on remote monitoring and video processing. By real-time online monitoring of the daily physical health status of all elderly people in the community, the management level and quality of community home elderly care services are improved, thereby enhancing the comprehensive elderly care experience of the elderly.

[0004] However, in actual applications, it is found that most of the remote home health care solutions in the prior art are based on a single dimension of data for limited threshold judgment and warning, and the monitoring accuracy is not high enough; and if more types / numbers of physical sensors are needed to improve the comprehensiveness of the data, the hardware cost and data energy consumption will be increased invisibly; and in the monitoring solutions of the prior art involving multiple types of sensors, the data of different dimensions are not linked, the data processing dimension is single and the processing method of each dimension is relatively complicated, which cannot reduce the data processing cost. Summary of the invention

[0005] In response to the above technical problems, the present invention proposes a remote multi-dimensional monitoring system and method for home health status, a computer-readable storage medium, a computer program product and an electronic device for implementing the method.

[0006] In a first aspect of the present invention, a remote home health status multi-dimensional monitoring system is provided, the system comprising a data acquisition layer, a data control layer, a data analysis layer and a data output layer;

[0007] The data acquisition layer includes image sensors, sound sensors and wearable devices;

[0008] The image sensor is used to collect image frames of the target person in N monitoring areas;

[0009] The sound sensor is used to collect the voiceprint signals existing in the N monitoring areas;

[0010] The target person wears the wearable device and moves in the N monitoring areas; N≥2;

[0011] The data control layer includes a first data queue corresponding to the image sensor and a second data queue corresponding to the sound sensor;

[0012] The data analysis layer analyzes at least part of the sensor data stored in the first data queue and / or the second data queue based on the control instruction generated by the data control layer;

[0013] The data output layer receives the analysis results of the data analysis layer, outputs the remote home health status of the target person, and sends a feedback indication to the wearable device worn by the target person when the remote home health status of the target person meets the preset standard.

[0014] The first data queue is a circular queue of size N; each monitoring area is configured with one image sensor;

[0015] At every preset period, the image frames corresponding to each monitoring area acquired by each image sensor are stored in the ring queue;

[0016] When the circular queue is full, the data analysis layer generates the control instruction to detect and analyze all the sensor data stored in the first data queue.

[0017] At every preset period, the image frames corresponding to each monitoring area collected by each image sensor are stored in the ring queue, specifically including:

[0018] At every preset period, N image frames corresponding to N monitoring areas acquired by N image sensors are randomly stored in the ring queue;

[0019] When the circular queue is full, the data analysis layer generates the control instruction to detect and analyze all the sensor data stored in the first data queue, specifically including:

[0020] Detect whether the target person appears in the N image frames.

[0021] Each monitoring area is configured with one sound sensor; at every preset period, the voiceprint signal corresponding to each monitoring area collected by each sound sensor is stored in the second data queue;

[0022] Furthermore, the i-th monitoring area corresponds to the voiceprint signal Vi and the image frame Fi; the second data queue includes N storage locations, and the voiceprint signal Vi is stored in the i-th storage location of the second data queue; i=1, 2, ..., N;

[0023] When the circular queue is full, the data analysis layer detects N image frames stored in the circular queue;

[0024] When the target person is not detected in the N image frames, at least part of the sensing data stored in the second data queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second data queue is continuous.

[0025] The remote home health status of the target person includes one of the following:

[0026] (1) The target person is not in any surveillance area;

[0027] (2) The target person is in one of the surveillance areas and remains relatively still for a long time and is outside the surveillance range;

[0028] (3) The target person moves normally in multiple surveillance areas.

[0029] When the remote home health status of the target person meets the preset standard, a feedback indication is sent to the wearable device worn by the target person, specifically including:

[0030] When the analysis result indicates that the target person is in a relatively static state for a long time and is outside the monitoring range, a feedback indication is sent to the wearable device worn by the target person, and the feedback indication requires the target person to make a feedback action or submit a voice response.

[0031] In a second aspect of the present invention, a method for remote multi-dimensional monitoring of health status at home is proposed.

[0032] The embodiments of the multi-dimensional monitoring method of home health status mentioned in this section can be implemented through a computer program on an electronic device or system configured with a memory and a processor. The electronic device or system can be in the form of a physical machine, a virtual machine, a server, a cluster, or any combination thereof.

[0033] The specific form of the electronic device may also be a human-computer interaction terminal, and the human-computer interaction terminal may be a desktop terminal, a smart handheld terminal, a mobile terminal, etc. with a human-computer interaction interface.

[0034] The method comprises the following steps:

[0035] S710: Collect N image frames of the target person in N monitoring areas and N groups of voiceprint signals existing in the N monitoring areas; N≥2;

[0036] S720: storing the N image frames into a first circular queue, and storing the N groups of voiceprint signals into a second common queue;

[0037] S730: When the first ring queue is full, analyzing at least part of the sensing data stored in the first ring queue and / or the second common queue;

[0038] S740: Based on the analysis result, output the remote home health status of the target person;

[0039] And when the remote home health status of the target person meets the preset standard, a feedback instruction is sent to the target person, and the feedback instruction requires the target person to make a feedback action or submit a voice reply.

[0040] Furthermore, the step S710 and the step S720 are repeatedly executed every preset period;

[0041] Furthermore, in different cycles, the step 720 stores the N image frames into the first circular queue in different orders;

[0042] The step 720 in different cycles stores the N groups of voiceprint signals into the second common queue in the same order.

[0043] When the first circular queue is full, detecting whether the target person appears in the N image frames of the first circular queue;

[0044] When the target person is not detected in the N image frames, at least part of the sensing data stored in the second general queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second general queue is continuous.

[0045] The aforementioned remote multi-dimensional monitoring method of home health status can be automatically implemented through various forms of electronic devices through computer program instructions; the computer program instructions can be stored in different forms of storage media and loaded into computer electronic devices for execution.

[0046] Therefore, in the third aspect of the present invention, a computer-readable storage medium is also provided for storing computer instructions. When the computer instructions are executed on an electronic device, the electronic device executes all or part of the steps of the aforementioned remote home health status multi-dimensional monitoring method.

[0047] In the fourth aspect of the present invention, a computer device is also proposed, which includes a processor and a memory, the memory is used to store instructions, and the processor is used to call the instructions in the memory, so that the computer device executes the aforementioned remote home health status multi-dimensional monitoring method.

[0048] In the fifth aspect of the present invention, a computer program product is also proposed, which includes a computer program. When the computer program is executed, all or part of the steps of the above-mentioned remote home health status multi-dimensional monitoring method are implemented.

[0049] The technical solution of the present invention collects and stores data respectively through a data acquisition layer composed of an image sensor, a sound sensor and a wearable device, a first data queue corresponding to the image sensor, and a data control layer composed of a second data queue corresponding to the sound sensor; the data analysis layer analyzes at least part of the sensor data stored in the first data queue and / or the second data queue based on the control instructions generated by the data control layer; the data output layer receives the analysis result of the data analysis layer, outputs the remote home health status of the target person, and sends a feedback indication to the wearable device worn by the target person when the remote home health status of the target person meets the preset standard, so that the remote home health status can be comprehensively determined and monitored based on multi-dimensional data, and different dimensions complement and activate each other, which can ensure accuracy while reducing sensor data energy consumption.

[0050] Further advantages of the present invention will be further reflected in detail in the specific embodiments section in conjunction with the drawings of the specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 It is a schematic diagram of the architecture hierarchy of a remote home health status multi-dimensional monitoring system according to an embodiment of the present invention;

[0053] Figure 2 It is a schematic diagram of some data interaction principles of the remote home health status multi-dimensional monitoring system of the present invention;

[0054] Figure 3 It is a schematic diagram of the layout of the target monitoring area and the data acquisition sensor in the technical solution of the present invention;

[0055] Figure 4 It is a flowchart of a method for multi-dimensional monitoring of remote home health status according to an embodiment of the present invention; DETAILED DESCRIPTION

[0056] See first Figure 1 , Figure 1 From the perspective of system architecture level, a schematic diagram of the architecture level of a remote home health status multi-dimensional monitoring system according to an embodiment of the present invention is introduced.

[0057] Figure 1 The remote home health status multi-dimensional monitoring system is generally shown to include a data acquisition layer, a data control layer, a data analysis layer and a data output layer.

[0058] The data acquisition layer includes image sensors, sound sensors and wearable devices;

[0059] The image sensor is used to collect image frames of the target person in N monitoring areas;

[0060] The sound sensor is used to collect the voiceprint signals existing in the N monitoring areas;

[0061] The target person wears the wearable device and moves in the N monitoring areas; N≥2;

[0062] The data control layer includes a first data queue corresponding to the image sensor and a second data queue corresponding to the sound sensor;

[0063] The data analysis layer analyzes at least part of the sensor data stored in the first data queue and / or the second data queue based on the control instruction generated by the data control layer;

[0064] The data output layer receives the analysis results of the data analysis layer, outputs the remote home health status of the target person, and sends a feedback indication to the wearable device worn by the target person when the remote home health status of the target person meets the preset standard.

[0065] Next, combine Figure 2 and Figure 3 , further introduces the functional implementation of the above-mentioned different levels of architecture.

[0066] exist Figure 2In the figure, it is shown that the data acquisition layer composed of an image sensor, a sound sensor and a wearable device, the first data queue corresponding to the image sensor, and the data control layer composed of a second data queue corresponding to the sound sensor respectively perform data acquisition and storage; the data analysis layer includes at least one data analysis engine, which analyzes at least part of the sensor data stored in the first data queue and / or the second data queue based on the control instructions generated by the data control layer; the data output layer receives the analysis results of the data analysis layer, outputs the remote home health status of the target person, and sends a feedback indication to the wearable device worn by the target person when the remote home health status of the target person meets the preset standard.

[0067] In the data control layer, the first data queue is a circular queue of size N, and the second data queue is a first-in-first-out common queue of size N. The value of N here is equal to the number of target monitoring areas.

[0068] by Figure 3 The target monitoring area and the layout diagram of the data acquisition sensor are shown as an example. Figure 3 , a home monitoring environment is shown, which is roughly divided into a living room (where image sensor F1 and sound sensor V1 are arranged), rooms (living rooms, kitchens and bathrooms, etc.) 1-4 (where image sensors F2-F5 and sound sensors V2-V5 are arranged accordingly).

[0069] For the convenience of description, in the subsequent embodiments, the data (image frames) collected by the image sensors F1-F5 are also described as F1-F5; the data (voiceprint sequences) collected by the sound sensors V1-V5 are also described as V1-V5;

[0070] As a first advantage of the present invention, Figure 3 In the five monitoring areas (N=5) shown, each monitoring area is configured with one image sensor and one sound sensor. That is, the technical solution of the present invention first needs to collect sensor data in two dimensions (image dimension and sound dimension).

[0071] However, next, the present application does not simply combine the sensor data of two dimensions (image dimension and sound dimension) for storage and processing, but makes the following improvements.

[0072] First, at every preset period, the image frames corresponding to each monitoring area collected by each image sensor are stored in the circular queue; when the circular queue is full, the data analysis layer generates the control instruction to detect and analyze all the sensor data stored in the first data queue.

[0073] The inventors noticed that when conducting remote home health status monitoring, the focus is on identifying abnormal conditions, but abnormal conditions are actually in the minority in the overall monitoring cycle. Despite this, uninterrupted sensor data collection is still required. Although the data collection process cannot be avoided, not every data collection has analysis value. Therefore, the present invention has made the above-mentioned improvements. First, the data storage method adopts a circular queue to avoid data overflow while ensuring that the latest data is analyzed each time; instead, analysis is performed only when the queue is full, avoiding frequent data analysis processes.

[0074] Secondly, in order to reflect the linkage of the above two processes, during the specific storage, every preset period, the N image frames corresponding to the N monitoring areas collected by the N image sensors are randomly stored in the circular queue. In other words, the N image frames are stored in the first circular queue in different orders in different periods.

[0075] The above process is described below with an illustrative example.

[0076] Assume that the preset period is 30s, N=5, t=0s is the starting point, and the ring queue storage size is 5 (all values ​​are only for example, and the values ​​in actual situations may be different).

[0077] Although the circular queue does not actually have a starting point and an end point (the circular storage covers and does not overflow), for the convenience of description, it is assumed that a certain storage area R1 is the starting point and the storage area R5 is the end point (R1 and R5 are connected end to end, that is, R1-R2-R3-R4-R5-R1-...);

[0078] t=0: the data (image frames) collected by image sensors F1-F5 are also described as F1-F5 accordingly;

[0079] Then F1 is stored in R1, F2 is stored in R3, F3 is stored in R2, F4 is stored in R5, and F5 is stored in R4.

[0080] t=30: the data (image frames) collected by the image sensors F1-F5 are also described as F'1-F'5 accordingly;

[0081] Then F'1 is stored in R3, F'2 is stored in R5, F'3 is stored in R'2, F'4 is stored in R4, and F'5 is stored in R1.

[0082] The key point of the above example is that the N image frames are randomly stored in the first circular queue in different orders in different cycles. The purpose of randomness is to interrupt the storage order each time, because the subsequent analysis also needs to consider the different order each time to improve accuracy. This is also in line with the random characteristics of the target person's activities between different monitoring areas. The prior art has not taken this feature into account and provided a corresponding data storage and processing method, which is one of the important improvements of the invention.

[0083] Correspondingly, when the ring queue is full, the data analysis layer generates the control instruction to detect and analyze all the sensor data stored in the first data queue, specifically including:

[0084] Detect whether the target person appears in the N image frames.

[0085] Preferably, considering the randomness of the storage order of the circular queue at each time, when the circular queue is full, the data analysis layer detects the N image frames stored in the circular queue;

[0086] When the target person is not detected in N consecutive image frames, at least part of the sensor data stored in the second data queue is analyzed.

[0087] If the target person is not detected in N consecutive image frames after a certain data collection and random order storage, there are two possibilities: one is that the target person is out and not in any monitoring area; the other is that the target person happens to be in a monitoring blind spot and has not moved for a long time. There are many reasons for this situation. Due to the limitations of the house structure and the limited sensors, no matter how the sensor position is set, there will always be some monitoring blind spots. Although these blind spots are the necessary channels connecting different areas, the target person usually does not stay for a long time. Therefore, considering the limited value of monitoring sensors, they will not be set specifically for this purpose.

[0088] However, if the target person is not detected in any of the N image frames, there is a high probability that an abnormality exists.

[0089] It can be seen that since a circular queue has been used to store N image frames, and the storage order of different cycles is random, the speed at which the detection process of "the target person has not been detected in N consecutive image frames" draws the conclusion is optimal as a whole, and there will be no error accumulation in extreme cases, thereby reducing the amount of data calculation while ensuring compliance with actual conditions.

[0090] For the first abnormality, additional monitoring and maintenance can be performed based on the positioning properties of the wearable device worn by the target person; however, for the second situation, further analysis is required, that is, when the target person is not detected in the N image frames, at least part of the sensor data stored in the second data queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second data queue is continuous.

[0091] Preferably, each monitoring area is equipped with one sound sensor; at every preset period, the voiceprint signal corresponding to each monitoring area collected by each sound sensor is stored in the second data queue;

[0092] Furthermore, the i-th monitoring area corresponds to the voiceprint signal Vi and the image frame Fi; the second data queue includes N storage locations, and the voiceprint signal Vi is stored in the i-th storage location of the second data queue; i=1, 2, ..., N;

[0093] When the circular queue is full, the data analysis layer detects N image frames stored in the circular queue;

[0094] When the target person is not detected in the N image frames, at least part of the sensing data stored in the second data queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second data queue is continuous.

[0095] It can be seen that we now enter the second dimension of data analysis.

[0096] Different from the storage method of image frame data in the first dimension, the voiceprint signal in the second dimension is stored in a fixed order to facilitate subsequent continuity analysis.

[0097] Continuing with the previous example, the explanation is as follows:

[0098] Assume that the preset period is 30s, N=5, t=0s is the starting point, and the storage size of the second data queue is 5 (all values ​​are examples and the actual situation may be different).

[0099] The second data queue is a first-in-first-out linked storage queue, and the storage positions are L1-L2-L3-L4-L5 in sequence; when 5 elements are stored, if there is a sixth element that needs to be queued, the first element stored will be removed from the queue.

[0100] t=0: the data collected by the sound sensors F1-F5 (voiceprint signal sequence) are also described as V1-V5 accordingly;

[0101] Then V1 is stored in L1, V2 is stored in L2, V3 is stored in L3, V4 is stored in L4, and V5 is stored in L5.

[0102] t=30: the data (image frames) collected by the sound sensors F1-F5 are also described as F'1-F'5 accordingly;

[0103] Then V'1 is stored in L1, V'2 is stored in L2, V'3 is stored in L3, V'4 is stored in L4, and V'5 is stored in L5.

[0104] At this time, when the target person is not detected in the N image frames, at least part of the sensing data stored in the second data queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second data queue is continuous.

[0105] That is to say, when the target person is not detected in the N image frames, it is only preliminarily determined that there may be some abnormal state, but it still needs to be further confirmed by the voiceprint signal.

[0106] Specifically, since the voiceprint signals are stored in the second data queue in a fixed order, if the voiceprint signals are discontinuous, it means that the target person has not passed through two consecutive different areas, that is, the target person may be in one of the monitoring areas and is in a relatively static state for a long time and is outside the monitoring range.

[0107] Preferably, the N storage areas included in the second data queue correspond to the N monitoring areas.

[0108] by Figure 3 For example, preferably, L1 corresponds to V1, L2 corresponds to V2, L3 corresponds to V4, L4 corresponds to V5, and L5 corresponds to V3.

[0109] That is, the continuous relationship of the N storage areas included in the second data queue is determined based on the continuous (adjacent) relationship of the N monitoring areas.

[0110] For example, if there is no voiceprint signal in L1, but there is a voiceprint signal in L2, and then there is no voiceprint signal in L3, it may mean that the target person is in the blind spot between areas V2-V4 ( Figure 3 Abnormal stop in the aisle.

[0111] Of course, the above examples of continuity or region proximity are merely illustrative, and those skilled in the art may set the continuity relationship based on the actual layout position of the target region and experience.

[0112] The improvement focus of the above embodiment is that based on the continuity relationship, the voiceprint signal is collected for chain queue storage and the health status of the monitoring area is judged based on the continuity relationship (that is, at least two storage areas), avoiding the impact of data loss in a single storage area.

[0113] The above monitoring and analysis are all carried out by Figure 3 The data analysis engine of the data analysis layer executes the analysis, and the analysis results are output by the data output layer.

[0114] Therefore, as a general example, the remote home health status of the target person includes one of the following:

[0115] (1) The target person is not in any surveillance area;

[0116] (2) The target person is in one of the surveillance areas and remains relatively still for a long time and is outside the surveillance range;

[0117] (3) The target person moves normally in multiple surveillance areas.

[0118] When the remote home health status of the target person meets the preset standard, a feedback indication is sent to the wearable device worn by the target person, specifically including:

[0119] When the analysis result indicates that the target person is in a relatively static state for a long time and is outside the monitoring range, a feedback indication is sent to the wearable device worn by the target person, and the feedback indication requires the target person to make a feedback action or submit a voice response.

[0120] Optionally, when the target person does not make a feedback action or submit a voice reply, it is determined that the target person is abnormal and relevant measures need to be taken, such as remotely calling for help.

[0121] Preferably, when the remote home health status of the target person meets the preset standard, a feedback indication is sent to the wearable device worn by the target person, specifically including:

[0122] When the analysis result indicates that the target person is not in any monitoring area, the wearable device worn by the target person is used to determine whether the target person is out. If so, a feedback indication is sent to the wearable device worn by the target person, and the feedback indication requires the target person to make a feedback action or submit a voice reply.

[0123] Preferably, when the remote home health status of the target person is that the target person is normally active in multiple monitoring areas, increasing the preset period can further realize the data collection energy consumption of the data sensor and the subsequent processing energy consumption.

[0124] exist Figure 1-Figure 3 Based on the system embodiment shown, Figure 4 A flow chart showing a method for multi-dimensional remote home health status monitoring according to an embodiment of the present invention.

[0125] exist Figure 4 In the method, the method can be implemented by a computer program on an electronic device or system configured with a memory and a processor. The electronic device or system can be in the form of a physical machine, a virtual machine, a server, a cluster, or any combination thereof.

[0126] The specific form of the electronic device may also be a human-computer interaction terminal, and the human-computer interaction terminal may be a desktop terminal, a smart handheld terminal, a mobile terminal, etc. with a human-computer interaction interface.

[0127] The method comprises the following steps:

[0128] S710: Collect N image frames of the target person in N monitoring areas and N groups of voiceprint signals existing in the N monitoring areas; N≥2;

[0129] S720: storing the N image frames into a first circular queue, and storing the N groups of voiceprint signals into a second common queue;

[0130] S730: When the first ring queue is full, analyzing at least part of the sensing data stored in the first ring queue and / or the second common queue;

[0131] S740: Based on the analysis result, output the remote home health status of the target person;

[0132] And when the remote home health status of the target person meets the preset standard, a feedback instruction is sent to the target person, and the feedback instruction requires the target person to make a feedback action or submit a voice reply.

[0133] Furthermore, the step S710 and the step S720 are repeatedly executed every preset period;

[0134] Furthermore, in different cycles, the step 720 stores the N image frames into the first circular queue in different orders;

[0135] The step 720 in different cycles stores the N groups of voiceprint signals into the second common queue in the same order.

[0136] When the first circular queue is full, detecting whether the target person appears in the N image frames of the first circular queue;

[0137] When the target person is not detected in the N image frames, at least part of the sensing data stored in the second general queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second general queue is continuous.

[0138] It can be understood that the corresponding functional modules and method steps of the method embodiment and the system embodiment should correspond to each other in terms of execution process and execution principle, so they will not be repeated here.

[0139] For other technologies, principles, algorithms or models not elaborated in detail in this application, please refer to the prior art.

[0140] In the aforementioned embodiment section, the present invention provides multiple embodiments, each of which can constitute an independent technical solution and may contribute to the prior art and solve corresponding technical problems.

[0141] However, it should be pointed out that different embodiments can be combined with each other without violating logic; at the same time, each embodiment can solve at least one technical problem, but it is not required that each individual embodiment must solve multiple or all technical problems.

[0142] At the same time, in the specific implementation of this application, if user-related data is involved, when the embodiment of this application is applied to a specific product or technology, the user's permission or consent must be obtained, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0143] The technical solution of the present invention collects and stores data through a data acquisition layer composed of an image sensor, a sound sensor and a wearable device, a first data queue corresponding to the image sensor, and a data control layer composed of a second data queue corresponding to the sound sensor. The data analysis layer analyzes at least part of the sensor data stored in the first data queue and / or the second data queue based on the control instructions generated by the data control layer. The data output layer receives the analysis result of the data analysis layer and outputs the remote home health status of the target person.

[0144] At the same time, the first data queue adopts a circular queue, and the second data queue adopts a normal queue, which conforms to the data generation characteristics of the actual home environment and the needs of subsequent data analysis;

[0145] Finally, when the target person's remote home health status meets the preset standards, a feedback indication is sent to the wearable device worn by the target person, which can perform comprehensive judgment and monitoring of the remote home health status based on multi-dimensional data. Different dimensions complement and activate each other, which can ensure accuracy while reducing sensor data energy consumption.

[0146] The foregoing has shown and described the method embodiments and system of the present invention, but it is understandable to those skilled in the art that various changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A remote home health status multi-dimensional monitoring system, the system comprising a data acquisition layer, a data control layer, a data analysis layer and a data output layer; characterized in that: The data acquisition layer includes image sensors, sound sensors and wearable devices; The image sensor is used to collect image frames of the target person in N monitoring areas; The sound sensor is used to collect the voiceprint signals existing in the N monitoring areas; The target person wears the wearable device and moves in the N monitoring areas; N≥2; The data control layer includes a first data queue corresponding to the image sensor and a second data queue corresponding to the sound sensor; The data analysis layer analyzes at least part of the sensor data stored in the first data queue and / or the second data queue based on the control instruction generated by the data control layer; The data output layer receives the analysis results of the data analysis layer, outputs the remote home health status of the target person, and sends a feedback indication to the wearable device worn by the target person when the remote home health status of the target person meets the preset standard.

2. A remote home health status multi-dimensional monitoring system as claimed in claim 1, characterized in that: The first data queue is a circular queue of size N; each monitoring area is configured with one image sensor; At every preset period, the image frames corresponding to each monitoring area acquired by each image sensor are stored in the ring queue; When the circular queue is full, the data analysis layer generates the control instruction to detect and analyze all the sensor data stored in the first data queue.

3. A remote home health status multi-dimensional monitoring system as claimed in claim 2, characterized in that: in, At every preset period, the image frames corresponding to each monitoring area collected by each image sensor are stored in the ring queue, specifically including: At every preset period, N image frames corresponding to N monitoring areas acquired by N image sensors are randomly stored in the ring queue; When the circular queue is full, the data analysis layer generates the control instruction to detect and analyze all the sensor data stored in the first data queue, specifically including: Detect whether the target person appears in the N image frames.

4. A remote home health status multi-dimensional monitoring system as claimed in claim 2, characterized in that: Each monitoring area is configured with one sound sensor; at every preset period, the voiceprint signal corresponding to each monitoring area collected by each sound sensor is stored in the second data queue; Furthermore, the i-th monitoring area corresponds to the voiceprint signal Vi and the image frame Fi; the second data queue includes N storage locations, and the voiceprint signal Vi is stored in the i-th storage location of the second data queue; i=1, 2, ..., N; When the circular queue is full, the data analysis layer detects N image frames stored in the circular queue; When the target person is not detected in the N image frames, at least part of the sensing data stored in the second data queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second data queue is continuous.

5. A remote home health status multi-dimensional monitoring system as claimed in claim 3 or 4, characterized in that: Features: The remote home health status of the target person includes one of the following: (1) The target person is not in any surveillance area; (2) The target person is in one of the surveillance areas and remains relatively still for a long time and is outside the surveillance range; (3) The target person moves normally in multiple surveillance areas.

6. A remote home health status multi-dimensional monitoring system as claimed in claim 3 or 4, characterized in that: Features: When the remote home health status of the target person meets the preset standard, a feedback indication is sent to the wearable device worn by the target person, specifically including: When the analysis result indicates that the target person is in a relatively static state for a long time and is outside the monitoring range, a feedback indication is sent to the wearable device worn by the target person, and the feedback indication requires the target person to make a feedback action or submit a voice response.

7. A remote multi-dimensional monitoring method for home health status, characterized in that: The method comprises the following steps: S710: Collect N image frames of the target person in N monitoring areas and N groups of voiceprint signals existing in the N monitoring areas; N≥2; S720: storing the N image frames into a first circular queue, and storing the N groups of voiceprint signals into a second common queue; S730: When the first ring queue is full, analyzing at least part of the sensing data stored in the first ring queue and / or the second common queue; S740: Based on the analysis result, output the remote home health status of the target person; And when the remote home health status of the target person meets the preset standard, a feedback instruction is sent to the target person, and the feedback instruction requires the target person to make a feedback action or submit a voice reply.

8. A remote multi-dimensional monitoring method for home health status as claimed in claim 7, characterized in that: Repeating step S710 and step S720 every preset period; Furthermore, in different cycles, the step 720 stores the N image frames into the first circular queue in different orders; The step 720 in different cycles stores the N groups of voiceprint signals into the second common queue in the same order.

9. A remote multi-dimensional monitoring method for home health status as claimed in claim 7, characterized in that: The step S730 includes: When the first circular queue is full, detecting whether the target person appears in the N image frames of the first circular queue; When the target person is not detected in the N image frames, at least part of the sensing data stored in the second general queue is analyzed, and the analysis includes detecting whether the voiceprint signal stored in the second general queue is continuous.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of a remote multi-dimensional monitoring method for home health status as described in any one of claims 7 to 9 are implemented.

Images