Indoor person falling detection method, device and equipment and storage medium

Abstract

The invention relates to the technical field of signal processing, and discloses an indoor person falling detection method, device and equipment and a storage medium, and the method comprises the steps: collecting the temperature distribution data of a target person through an infrared array sensor disposed in a target indoor space; based on the temperature distribution data, transverse morphological characteristics and longitudinal distance characteristics of the target person are extracted, the transverse morphological characteristics are used for representing spatial changes of human body postures, and the longitudinal distance characteristics are used for analyzing dynamic evolution of action trends; and inputting the fused feature set into a classification model, and outputting a tumble state judgment result. According to the invention, non-contact data acquisition is realized through an infrared sensing technology, and the accuracy of fall detection is effectively improved in combination with transverse and longitudinal feature fusion and a classification algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and particularly relates to an indoor personnel fall detection method, device, equipment and storage medium. Background Art

[0002] The importance of indoor fall monitoring for life safety is becoming increasingly prominent. In the prior art, fall detection schemes are mainly divided into two categories: wearable and non-wearable. Wearable devices rely on sensors worn by the human body, and have problems such as inconvenient use and high false alarm rates; while non-wearable solutions usually adopt methods based on image or sensor data analysis, but are limited by environmental interference or data processing complexity, and it is difficult to achieve high-precision real-time monitoring.

[0003] Therefore, there is an urgent need for a non-contact fall detection method that can overcome the defects of the prior art and improve the accuracy of fall detection. Summary of the Invention

[0004] In view of this, the present invention provides an indoor personnel fall detection method, device, equipment and storage medium to improve the accuracy of fall detection. The technical solution is as follows.

[0005] In a first aspect, the present invention provides an indoor personnel fall detection method, which includes: Obtain the temperature distribution data of a target person in a target indoor space; the temperature distribution data is collected by an infrared array sensor deployed in the target indoor space; Based on the temperature distribution data, extract the lateral morphological features and longitudinal distance features of the target person to obtain a fusion feature set; the lateral morphological features are used to characterize the posture change of the target person; the longitudinal distance features are used to characterize the movement trend of the target person; Input the fusion feature set into a classification model to obtain a determination result of the fall state of the target person.

[0006] An indoor personnel fall detection method provided by the present invention collects temperature distribution data of target personnel in real time through an infrared array sensor deployed in the target indoor space, replacing traditional wearable devices or microwave radars with non-contact thermal imaging technology, avoiding the inconvenience of using sensors close to the body, and the infrared array sensor only collects low-resolution heat source distributions without requiring high-definition images or electromagnetic wave reflection signals, simplifying the complexity of hardware deployment. By extracting horizontal morphological features (reflecting human postures) and vertical distance features (reflecting action trends) and constructing a fused feature set, the composite characteristics of fall behaviors can be comprehensively captured from two dimensions of spatial distribution and temporal evolution. Compared with single-dimensional feature description methods (such as only analyzing speed or acceleration), the false judgment probability caused by blurred partial features is significantly reduced; further inputting the fused feature set into a classification model for intelligent determination can handle complex situations of fall behaviors, overcoming the problem of insufficient sensitivity caused by environmental interference or individual differences in traditional threshold rule methods, thereby improving the accuracy of indoor personnel fall detection under the premise of non-invasive detection.

[0007] In an alternative embodiment, before extracting the horizontal morphological features and vertical distance features of the target personnel, it further includes: Calculating the variance of the temperature distribution data of consecutive frames; When it is detected that the variance of the temperature distribution data of two or more consecutive frames exceeds the variance threshold, the horizontal morphological features and vertical distance features of the target personnel are extracted.

[0008] An indoor personnel fall detection method provided by the present invention, by setting a variance threshold, only starts feature extraction (such as violent human actions) when significant temperature changes are detected, avoiding false triggers caused by minor fluctuations in environmental temperature (such as air conditioner airflow, pet movement), and significantly reducing the system false alarm rate. Judging multiple consecutive frames (two or more frames) can filter transient noise interference (such as short-term heat source interference) and improve the reliability of feature extraction. Feature extraction is only performed when effective actions occur, reducing the amount of invalid data processing.

[0009] In an alternative embodiment, extracting the horizontal morphological features of the target personnel includes: Obtaining the number of temperature pixel points exceeding the preset room temperature threshold in each frame of temperature distribution data and comparing it with the total number of frames to obtain the average number of effective pixel points; Performing contour analysis on the active pixel points in consecutive activation frames to extract morphological features; Determining the effective action area of the human action of the target personnel based on the morphological features; Based on the average number of effective pixel points and the effective action area, obtaining the horizontal morphological features of the target personnel; Among them, the consecutive activation frames are the frame sequence between the starting frame and the ending frame; the starting frame is the frame where the variance of the temperature distribution data of two consecutive frames is first detected to exceed the variance threshold; the ending frame is the frame where the variance of the temperature distribution data of two consecutive frames is first detected to be lower than the variance threshold. An indoor personnel fall detection method provided by the present invention extracts the average number of effective pixel points and the effective action area. The average number of effective pixel points reflects the persistence and spatial coverage of the action, enhancing the ability to distinguish action types. The activation frames record the start and end times of the action, improving the recognition accuracy of the temporal features of the fall behavior. The effective action area approximates the pixel density and contour, mapping the human body posture change into geometric parameters, avoiding misjudgment caused by relying only on a single parameter (such as speed) in traditional methods, and enhancing the sensitivity to the fall amplitude.

[0010] In an alternative embodiment, extracting the longitudinal distance feature of the target person includes: According to the maximum value in the variance of the temperature distribution data of each frame, obtain the maximum temperature distribution variance; Between the start frame and the end frame when the target person makes a human movement, obtain the number of pixel points when the variance of the temperature distribution data of each frame is greater than the temperature difference threshold, and obtain the maximum number of reaction pixels; Based on the maximum temperature distribution variance and the maximum number of reaction pixels, obtain the longitudinal distance feature of the target person.

[0011] An indoor personnel fall detection method provided by the present invention, the maximum temperature variance directly reflects the severity of the temperature change, improving the detection sensitivity to sudden actions. The maximum number of pixels characterizes the area size affected by the action. Combining the maximum temperature variance can distinguish falls from other large-scale actions (such as jumping), reducing false alarms. Solve the problem that single-dimensional features (such as only speed or acceleration) in traditional methods are difficult to describe complex fall behaviors.

[0012] In an alternative embodiment, inputting the fusion feature set into a classification model to obtain the determination result of the fall state of the target person, including: Assign different weight factors to each feature in the fusion feature set to obtain fusion features with different weights; Input the fusion features with different weights into the classification model to obtain the determination result of the fall state of the target person.

[0013] An indoor personnel fall detection method provided by the present invention, the weight factor is dynamically adjusted according to the contribution degree of the feature to fall recognition, solving the classification deviation problem caused by equal-weight processing of all features in traditional detection methods, and improving the attention of the model to key features. Through weight adjustment, it can adapt to different environments or individual differences, enhancing the generalization of the algorithm.

[0014] In an optional implementation, the classification model includes a K-nearest neighbor algorithm model or a neural network model.

[0015] The present invention provides a method for detecting indoor falls. In small sample scenarios (such as limited home user data), KNN quickly identifies fall patterns through similarity measurement, avoiding the limitation that neural networks require a large amount of training data and lowering the deployment threshold. In big data scenarios (such as multi-user data in nursing homes), neural networks capture nonlinear patterns of fall behaviors (such as implicit associations between different fall postures) through deep feature extraction, improving detection accuracy in complex scenarios. The model type is selected according to the application scenario requirements (such as KNN when real-time requirements are high, and neural network when accuracy requirements are high) to meet diverse user needs.

[0016] In an optional implementation, the infrared array sensor includes a top-view sensor group deployed on the top of the target indoor space and a side-view sensor group deployed on the side of the target indoor space.

[0017] The present invention provides a method for detecting indoor falls of people. The top sensor captures the vertical posture changes of the human body (such as a sudden drop in height when falling), the side sensor monitors the horizontal displacement (such as the lateral fall trajectory), and the multi-view fusion eliminates the blind spots of single sensors (such as the difficulty of the top-view sensor to detect the ground-hugging action). Through the joint analysis of the top-view and side-view data, the two-dimensional temperature distribution is mapped into a three-dimensional spatial motion trajectory (such as the fall angle, the contact point position), and the geometric description capability of the fall posture is enhanced. Multiple sensors can reduce the risk of detection failure caused by the obstruction of a single sensor (such as the furniture obstructing the side-view sensor), and improve the reliability of the system.

[0018] In summary, the indoor personnel fall detection method provided by the present invention forms a spatial complementary effect through the coordinated deployment of multiple groups of infrared array sensor groups in the data collection stage. The top-view sensor captures the sudden drop in height in the vertical direction (such as the downward shift of the center of gravity of the human body when falling), and the side-view sensor monitors the horizontal displacement trajectory (such as the lateral fall path). The multi-view data fusion provides a three-dimensional spatial modeling basis for the subsequent extraction of lateral morphological features and longitudinal distance features, which solves the problem of missed detection caused by the blind spot of the single sensor perspective; in the feature extraction stage, the variance threshold trigger mechanism and the fusion of horizontal and vertical features form a time-series-space association: the feature extraction is started by the continuous frame temperature variance exceeding the threshold, and the spatiotemporal parameters of the effective action area and the maximum temperature variance are linked to realize the recognition of the action starting point and the precise segmentation of the dynamic process, avoiding the defect of misjudgment of a single time point in the traditional scheme; in the classification stage, the weight factor allocation and the selectivity of the classification model form a dynamic adaptation closed loop, the lateral morphological features and the longitudinal distance features dynamically adjust the classification boundary through the contribution weight, and the elastic switching of the KNN model (small sample scenario) and the neural network (big data scenario) is combined to realize the generalization adaptation of the algorithm to different scenarios.

[0019] In a second aspect, the present invention provides an indoor personnel fall detection device, which includes: An acquisition module, configured to acquire temperature distribution data of a target person in a target indoor space; the temperature distribution data is collected by an infrared array sensor deployed in the target indoor space; An extraction module, configured to extract the lateral morphological features and longitudinal distance features of the target person based on the temperature distribution data to obtain a fusion feature set; the lateral morphological features are used to characterize the posture change of the target person; the longitudinal distance features are used to characterize the movement trend of the target person; A judgment module, configured to input the fusion feature set into a classification model to obtain a judgment result of the fall state of the target person.

[0020] In a third aspect, the present invention provides a computer device, including: a memory and a processor, which are communicatively connected to each other, and computer instructions are stored in the memory. The processor executes the computer instructions to execute the indoor personnel fall detection method according to the first aspect or any corresponding embodiment thereof.

[0021] In a fourth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to cause a computer to execute the indoor personnel fall detection method according to the first aspect or any corresponding embodiment thereof.

[0022] In a fifth aspect, the present invention provides a computer program product, including computer instructions, and the computer instructions are used to cause a computer to execute the indoor personnel fall detection method according to the first aspect or any corresponding embodiment thereof. Description of the Drawings

[0023] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0024] Figure 1 is a schematic flowchart of an indoor personnel fall detection method according to an embodiment of the present invention; Figure 2 is a schematic flowchart of the indoor personnel fall detection method of a specific example of the present invention; Figure 3 is a structural block diagram of an indoor personnel fall detection device according to an embodiment of the present invention; Figure 4 is a schematic hardware structure diagram of a computer device according to an embodiment of the present invention. Detailed implementation manners

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] To overcome the defects of existing fall detection solutions, an embodiment of a method for detecting the fall of indoor personnel is provided in an embodiment of the present invention. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0027] The method flow of the method for detecting the fall of indoor personnel in this embodiment is as Figure 1 shown and includes the following steps: S101. Obtain the temperature distribution data of a target person in a target indoor space; the temperature distribution data is collected by an infrared array sensor deployed in the target indoor space.

[0028] Specifically, the thermal radiation signal of the target person is collected in real time by an infrared array sensor (non-contact thermal imaging device) deployed in the target indoor space to generate a two-dimensional temperature distribution matrix. Each matrix unit (pixel point) corresponds to the temperature value of a certain position in the indoor space, forming dynamic data reflecting the thermal distribution on the human body surface. Only relying on infrared thermal imaging technology, without involving other sensing means such as visible light cameras or microwave radars; the temperature distribution data is a low-resolution thermal map (such as an 8×8 pixel array) to avoid privacy issues that may be caused by high-definition imaging.

[0029] S102. Based on the temperature distribution data, extract the horizontal morphological features and vertical distance features of the target person to obtain a fusion feature set; the horizontal morphological features are used to characterize the posture changes of the target person; the vertical distance features are used to characterize the movement trend of the target person.

[0030] Specifically, the horizontal morphological feature is to extract parameters related to the spatial distribution of the human body posture from the temperature distribution data, and this feature is used to quantify the morphological changes of the human body posture in the horizontal direction (such as standing upright, tilting, falling to the ground). For example, the geometric shape of the human body contour (such as the approximate rectangular area, aspect ratio), and the spatial aggregation degree of the effective temperature pixel points (such as the centroid position offset). The longitudinal distance feature is to extract parameters related to the evolution of the action time sequence from the temperature distribution data, and this feature is used to characterize the dynamic trend of the human body action in the time dimension (such as the speed difference between sitting down slowly and falling quickly). For example, the moving speed of the center point of the temperature distribution; the heat area diffusion rate (such as the sudden change of the local temperature field caused by the human body touching the ground when falling). The horizontal morphological feature (spatial distribution parameter) and the longitudinal distance feature (time sequence dynamic parameter) are combined into a multi-dimensional vector to form a fused feature set.

[0031] S103. Input the fused feature set into the classification model to obtain the determination result of the target person's fall state.

[0032] Specifically, the classification model constructed based on machine learning algorithms (such as KNN, SVM, or neural network) receives the fused feature set as input, learns the feature differences between normal activities and fall behaviors through training, and outputs a binary classification result (fall / non-fall). The classification model calculates the probability according to the distribution position of the fused feature set in the multi-dimensional feature space (such as the distance from the fall cluster), and if it exceeds the preset threshold, it is determined as a fall. For example, when the contour area suddenly increases (horizontal morphology) and the speed exceeds the critical value (longitudinal distance), the model determines it as a fall.

[0033] Optionally, in the above step S101, the infrared array sensor includes an overhead sensor group deployed on the top of the target indoor space and a side-view sensor group deployed on the side of the target indoor space. The overhead sensors capture the vertical posture changes of the human body (such as the sudden drop in height when falling), and the side-view sensors monitor the horizontal displacement (such as the lateral fall trajectory). The multi-view fusion eliminates the blind spots of a single sensor (such as it is difficult for the overhead sensor to detect the action of touching the ground). Through the joint analysis of the overhead and side-view data, the two-dimensional temperature distribution is mapped into a three-dimensional space action trajectory (such as the fall angle, contact point position), enhancing the geometric description ability of the fall posture. The multi-sensors can reduce the risk of detection failure caused by a single sensor being blocked (such as the side-view sensor being blocked by furniture), improving the reliability of the system.

[0034] Optionally, before the above step S102, it also includes a judgment on whether to perform feature extraction, specifically including: calculating the variance of the temperature distribution data of consecutive frames; when it is detected that the variance of the temperature distribution data of two or more consecutive frames exceeds the variance threshold, extracting the lateral morphological features and longitudinal distance features of the target person. By setting the variance threshold, feature extraction is only initiated when significant temperature changes are detected (such as violent human movements), avoiding false triggers caused by minor fluctuations in the ambient temperature (such as air conditioner airflow, pet movement), and significantly reducing the system false alarm rate. Judging multiple consecutive frames (two or more) can filter out transient noise interference (such as short-term heat source interference) and improve the reliability of feature extraction. Feature extraction is only performed when effective actions occur, reducing the amount of invalid data processing.

[0035] Optionally, the above step S102 includes the extraction of lateral morphological features and longitudinal morphological features.

[0036] Extracting the lateral morphological features specifically includes: obtaining the number of temperature pixel points exceeding the preset room temperature threshold in each frame of temperature distribution data and comparing it with the total number of frames to obtain the average number of effective pixel points; performing contour analysis on the active pixel points in consecutive activation frames to extract morphological features; determining the effective action area of the target person's body movement based on the morphological features; obtaining the lateral morphological features of the target person based on the average number of effective pixel points and the effective action area; where the consecutive activation frames are the frame sequence between the start frame and the end frame; the start frame is the frame where it is first detected that the variance of the temperature distribution data of two consecutive frames exceeds the variance threshold; the end frame is the frame where it is first detected that the variance of the temperature distribution data of two consecutive frames is lower than the variance threshold.

[0037] In the above steps, the extraction of the effective action area includes the body contour area and the centroid offset. Among them, the contour area is obtained by binarizing the temperature distribution matrix and counting the number of effective pixel points, and the centroid offset is calculated by the distance between the geometric center of the effective pixel point area and the reference position. The contour area is obtained through the contour analysis method. Contour analysis is to binarize the temperature distribution matrix, extract the effective pixel point area, and calculate the human projection area using the number of pixel points and the actual area of a unit pixel.

[0038] Extracting the longitudinal distance feature includes the moving speed of the center point of the temperature distribution. The moving speed is calculated by the ratio of the distance between the center points of adjacent frames within a sliding time window to the time interval and is processed by moving average filtering. Specifically, it includes: obtaining the maximum temperature distribution variance according to the maximum value in the variance of each frame of temperature distribution data; between the start frame and the end frame when the target person makes a body movement, obtaining the number of pixel points when the variance of the temperature distribution data of each frame is greater than the temperature difference threshold to obtain the maximum number of pixels; obtaining the longitudinal distance feature of the target person based on the maximum temperature distribution variance and the maximum number of pixels reflecting the pixels.

[0039] Among the above-extracted features, the average number of effective pixel points is defined as the total number of temperature pixel points exceeding the room temperature threshold divided by the total number of frames, that is, the average number of pixel points per frame exceeding the room temperature threshold; the activation frame represents the duration during which an action occurs. If there is no one in the indoor environment, this feature value is very small and basically does not change, but when someone is present and moving, this feature value will increase; the maximum number of responsive pixels represents the maximum number of pixels when the variance of the number of single pixels per frame is greater than the temperature difference threshold between the start frame and the end frame of human movement. This feature represents the number of pixel points collected during the corresponding action under the infrared sensor; the maximum temperature distribution variance represents the maximum temperature distribution variance between the start frame and the end frame of human movement. This feature represents the degree of change in the movement trend; the effective action area is to approximately outline the approximate contour of the human body's action into a regular rectangle, and calculate the approximate area of the action as the effective action area. By measuring the number of active pixel points in each frame of the activation frame divided by the number of pixel points per square, the approximate area where the actual action occurs in this frame can be obtained. This data feature can not only be used as a training feature, but also as verification data to verify whether the algorithm is accurate. By comparing the effective action area with the actual estimated action area, it can be determined which action has occurred.

[0040] Optionally, in the above step S103, different weight factors are first assigned to each feature in the fusion feature set to obtain fusion features with different weights; then the fusion features with different weights are input into the classification model to obtain the determination result of the target person's fall state. The weight factors are dynamically adjusted according to the contribution degree of the features to fall recognition, solving the classification deviation problem caused by equal-weight processing of all features in traditional detection methods and enhancing the model's attention to key features. Through weight adjustment, it can adapt to different environments or individual differences and enhance the generalization of the algorithm.

[0041] In the above steps, the classification model can adopt the K-nearest neighbor algorithm model or the neural network model. In a small-sample scenario (such as limited household user data), KNN can quickly identify fall patterns through similarity measurement, avoiding the limitation of the large amount of training data required by the neural network and reducing the deployment threshold. In a large-data scenario (such as multi-user data in a nursing home), the neural network can capture the non-linear patterns of fall behavior (such as the implicit associations of different fall postures) through deep feature extraction, improving the detection accuracy in complex scenarios. Select the model type according to the application scenario requirements (such as selecting KNN when high real-time performance is required and selecting the neural network when high accuracy is required) to meet the diverse user needs.

[0042] In summary, for the indoor personnel fall detection method provided by the embodiments of the present invention, in the data collection stage, through the collaborative deployment of multiple infrared array sensor groups, a spatial complementary effect is formed. The top-down sensors capture the sudden drop in height in the vertical direction (such as the downward movement of the human body's center of gravity during a fall), and the side-looking sensors monitor the horizontal displacement trajectory (such as the lateral fall path). The multi-perspective data fusion provides a three-dimensional space modeling basis for the subsequent extraction of lateral morphological features and longitudinal distance features, solving the problem of missed detection caused by the blind spot of a single sensor's perspective; in the feature extraction stage, the variance threshold trigger mechanism and the horizontal and vertical feature fusion form a time-sequence - space correlation: the feature extraction is started by the temperature variance of consecutive frames exceeding the threshold, and the spatio-temporal parameters of the effective action area and the maximum temperature variance are combined for linkage, realizing the recognition of the action starting point and the precise segmentation of the dynamic process, avoiding the defect of misjudgment at a single time point in the traditional scheme; in the classification stage, the weight factor allocation and the selectability of the classification model form a dynamic adaptation closed-loop. The lateral morphological features and the longitudinal distance features dynamically adjust the classification boundary through the contribution degree weight, and the elastic switching between the KNN model (small sample scenario) and the neural network (big data scenario) is combined to realize the generalization adaptation of the algorithm to different scenarios.

[0043] Exemplarily, a specific example is used below to illustrate the process of the above indoor personnel fall detection method. The method process of this example is as Figure 2 shown and includes the following steps.

[0044] First, after obtaining each frame of array data through the infrared array sensors deployed on the top and side of the indoor, perform the maximum variance operation on the temperature pixel points. If the variance of the temperature distribution of two or more consecutive frames is greater than the temperature variance threshold, start to extract the sensitive data features of the human body posture, including features such as the average number of effective pixel points, activation frames, the maximum number of responsive pixels, the maximum temperature distribution variance, and the effective action area. Finally, preprocess the extracted feature data according to the weight factor and then perform KNN classification to determine whether a fall action occurs. In addition to the KNN algorithm, other learning algorithms with the same classification effect, such as the random forest algorithm, can also be used.

[0045] Adopt infrared technology with the target human body as the source to replace microwave radar technology, use a low-resolution camera with strong privacy to collect human body information in real time, use multi-dimensional parameter fusion means to extract features, and use artificial intelligence algorithms to make recognition and determination of the extracted features, so as to realize the precise fall monitoring of the target human body. The contribution degrees of the features in each dimension to fall recognition are different. In the feature space of the classification algorithm, when measuring the distance between each feature parameter and the nearest neighbor, a weight factor corresponding to the fall contribution degree is given, thereby completing the adaptive optimization of the classification algorithm model in the multi-dimensional fusion scenario.

[0046] In this embodiment, an indoor personnel fall detection device is further provided. This device is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be elaborated again. As used hereinafter, the term "module" can be a combination of software and / or hardware that realizes a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0047] The structure of an indoor personnel fall detection device provided in this embodiment is as Figure 3 shown and includes: An acquisition module 301, configured to acquire temperature distribution data of a target person in a target indoor space; the temperature distribution data is collected by an infrared array sensor deployed in the target indoor space; An extraction module 302, configured to extract the lateral morphological features and longitudinal distance features of the target person based on the temperature distribution data to obtain a fusion feature set; the lateral morphological features are used to characterize the posture change of the target person; the longitudinal distance features are used to characterize the movement trend of the target person; A judgment module 303, configured to input the fusion feature set into a classification model to obtain a determination result of the fall state of the target person.

[0048] In an alternative embodiment, the device further includes: A threshold verification module 304, configured to calculate the variance of the temperature distribution data of consecutive frames; when it is detected that the variance of the temperature distribution data of two or more consecutive frames exceeds the variance threshold, extract the lateral morphological features and longitudinal distance features of the target person.

[0049] In an alternative embodiment, the extraction module 302 is specifically configured to: Obtain the number of temperature pixel points exceeding a preset room temperature threshold in each frame of temperature distribution data, and compare it with the total number of frames to obtain the average number of effective pixel points; Perform contour analysis on the active pixel points in consecutive activation frames to extract morphological features; Determine the effective action area of the human action of the target person based on the morphological features; Based on the average number of effective pixel points and the effective action area, obtain the lateral morphological features of the target person; Wherein, the consecutive activation frames are a frame sequence between the start frame and the end frame; the start frame is the frame when it is first detected that the variance of the temperature distribution data of two consecutive frames exceeds the variance threshold; the end frame is the frame when it is first detected that the variance of the temperature distribution data of two consecutive frames is lower than the variance threshold. In an alternative embodiment, the extraction module 302 is specifically configured to: Obtain the maximum temperature distribution variance according to the maximum value in the variance of each frame of temperature distribution data; Between the start frame and the end frame of the human movement of the target person, obtain the number of pixel points when the variance of the temperature distribution data of each frame is greater than the temperature difference threshold, and obtain the maximum number of pixels; Based on the maximum temperature distribution variance and the maximum number of reaction pixels, obtain the longitudinal distance feature of the target person.

[0050] In an alternative embodiment, the determination module 303 is specifically configured to: Assign different weight factors to each feature in the fusion feature set to obtain fusion features with different weights; Input the fusion features with different weights into the classification model to obtain the determination result of the falling state of the target person.

[0051] In an alternative embodiment, the classification model includes a K-nearest neighbor algorithm model or a neural network model.

[0052] The further function descriptions of the above-mentioned various modules and units are the same as those in the corresponding embodiments above, and will not be repeated here.

[0053] The indoor personnel fall detection device in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0054] The embodiment of the present invention also provides a computer device having the above-mentioned Figure 3 indoor personnel fall detection device.

[0055] Please refer to Figure 4 , Figure 4 which is a schematic structural diagram of a computer device provided by an alternative embodiment of the present invention. As Figure 4 shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some alternative embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (such as a server array, a set of blade servers, or a multi-processor system).Figure 4 Take a processor 10 as an example.

[0056] The processor 10 can be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 can further include a hardware chip. The above-mentioned hardware chip can be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device can be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.

[0057] Among them, the memory 20 stores instructions that can be executed by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiments.

[0058] The memory 20 can include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the computer device, etc. In addition, the memory 20 can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 can optionally include a memory remotely provided with respect to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and a combination thereof.

[0059] The memory 20 can include a volatile memory, such as a random access memory; the memory can also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 can also include a combination of the above types of memories.

[0060] The computer device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 can be connected through a bus or other means. Figure 4 Take the connection through a bus as an example.

[0061] The input device 30 can receive input digital or character information, and generate key signal inputs related to the user settings and function controls of the computer device, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 can include a display device, an auxiliary lighting device (such as an LED), and a tactile feedback device (such as a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light-emitting diode, a display, and a plasma display. In some alternative embodiments, the display device can be a touch screen.

[0062] Embodiments of the present invention also provide a computer-readable storage medium. The method according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code that is originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiments is implemented.

[0063] A part of the present invention can be applied as a computer program product, for example, computer program instructions. When executed by a computer, through the operation of the computer, the methods and / or technical solutions according to the present invention can be invoked or provided. Those skilled in the art should be able to understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Herein, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.

[0064] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for detecting indoor person falling, characterized in that: The method comprises: Acquire temperature distribution data of a target person in a target indoor space; the temperature distribution data is collected by an infrared array sensor deployed in the target indoor space; Based on the temperature distribution data, the transverse morphological features and the longitudinal distance features of the target person are extracted to obtain a fusion feature set; the transverse morphological features are used to characterize the posture changes of the target person; and the longitudinal distance features are used to characterize the movement trend of the target person; The fused feature set is input into a classification model to obtain a fall status determination result of the target person.

2. The method according to claim 1, characterized in that Before extracting the horizontal shape feature and the vertical distance feature of the target person, the method further includes: Calculate the variance of temperature distribution data of consecutive frames; When it is detected that the variance of temperature distribution data of two or more consecutive frames exceeds the variance threshold, the lateral morphological features and longitudinal distance features of the target person are extracted.

3. The method according to claim 2, characterized in that Extract the horizontal morphological features of the target person, including: Obtain the number of temperature pixels exceeding the preset room temperature threshold in each frame of temperature distribution data, and compare it with the total number of frames to obtain the average number of valid pixels; Perform contour analysis on active pixels in consecutive activation frames to extract morphological features; Determine the effective action area of ​​the target person's human body action based on the morphological characteristics; Based on the average number of effective pixels and the effective action area, a lateral morphological feature of the target person is obtained; Among them, the continuous activation frame is a frame sequence between a start frame and an end frame; the start frame is a frame where the variance of two consecutive frames of temperature distribution data is detected for the first time to exceed the variance threshold; the end frame is a frame where the variance of two consecutive frames of temperature distribution data is detected for the first time to be lower than the variance threshold.

4. The method according to claim 3, characterized in that Extract the longitudinal distance features of the target person, including: According to the maximum value of the variance of each frame of temperature distribution data, the maximum temperature distribution variance is obtained; Between the start frame and the end frame of the target person's body movement, the number of pixel points when the variance of the temperature distribution data of each frame is greater than the temperature difference threshold is obtained to obtain the maximum number of reaction pixels; Based on the maximum temperature distribution variance and the maximum number of reaction pixels, the longitudinal distance feature of the target person is obtained.

5. The method according to claim 4, characterized in that The step of inputting the fused feature set into a classification model to obtain a fall state determination result of the target person includes: Assigning different weight factors to each feature in the fusion feature set to obtain fusion features with different weights; The fusion features with different weights are input into the classification model to obtain the fall status determination result of the target person.

6. The method according to claim 5, characterized in that The classification model includes a K-nearest neighbor algorithm model or a neural network model.

7. The method according to any one of claims 1 to 6, characterized in that: The infrared array sensor comprises a top-view sensor group disposed on the top of the target indoor space and a side-view sensor group disposed on the side of the target indoor space.

8. An indoor person fall detection device, characterized in that: The device comprises: An acquisition module is used to acquire temperature distribution data of a target person in a target indoor space; the temperature distribution data is acquired by an infrared array sensor deployed in the target indoor space; An extraction module, used to extract the transverse morphological features and longitudinal distance features of the target person based on the temperature distribution data to obtain a fusion feature set; the transverse morphological features are used to characterize the posture changes of the target person; the longitudinal distance features are used to characterize the movement trend of the target person; The judgment module is used to input the fused feature set into the classification model to obtain the fall status judgment result of the target person.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the indoor person fall detection method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the indoor person fall detection method according to any one of claims 1 to 7.

Images