Infant side sleep identification method, device and equipment based on thermal infrared data analysis

Abstract

The invention relates to the technical field of infant nursing, solves the problem that the side sleep posture of an infant cannot be accurately recognized in the prior art, and provides an infant side sleep recognition method, device and equipment based on thermal infrared data analysis. The method comprises the following steps: acquiring temperature values of pixel points in an infant head area in a thermal infrared real-time image in an infant nursing scene; according to the temperature value, real-time statistical feature information corresponding to heat distribution in the head area of the infant is determined; calculating the real-time statistical feature information to obtain a skew coefficient of heat distribution; according to the deflection coefficient and a preset first threshold value, a preset second threshold value and a preset third threshold value corresponding to the side sleeping posture, the side sleeping posture of the infant is recognized, the side sleeping posture comprises light side sleeping, moderate side sleeping or severe side sleeping, the first threshold value is smaller than the second threshold value, and the second threshold value is smaller than the third threshold value. According to the invention, the accuracy of identifying the sleeping posture of the infant sleeping on one side is improved.

Description

[0001] This application is a divisional application of the invention patent application filed on May 10, 2024, with the invention name “Infant and young child sleeping position recognition method, device and equipment based on multimodal data fusion” and application number 202410572655.X. Technical Field

[0002] The present invention relates to the technical field of infant care, and in particular to a method, device and equipment for identifying side sleeping of infants based on thermal infrared data analysis. Background Art

[0003] Nowadays, in the care scene for infants and young children, detecting the sleeping position of infants and young children, especially distinguishing the degree of side sleeping, is very necessary to ensure the safety of infants and young children and promote their healthy development. Different degrees of side sleeping postures can increase the risk of suffocation in some cases, especially when infants and young children are on soft mattresses or close to loose bedding. Therefore, by accurately identifying the side sleeping posture of infants and young children, timely reminders can be issued to help caregivers or parents adjust the baby's sleeping position, reduce the risk of accidental suffocation, and ensure that infants and young children get enough sleep in a safe environment. The technology of accurate recognition of side sleeping posture is also of great significance for preventing sudden infant death syndrome (SIDS). The technology of monitoring the side sleeping posture of infants and young children has made significant progress in the past few years, including sensors embedded in cribs, wearable devices, smart cameras, and systems using artificial intelligence for real-time analysis. However, there are still some shortcomings. The existing technology cannot accurately identify all changes in side sleeping postures, especially in low light or blocked conditions, which can easily cause false alarms or missed alarms.

[0004] The existing Chinese patent CN115171033A discloses a method, device and equipment for intelligently monitoring the sleeping posture of infants and young children in combination with heterogeneous image sources. The method includes: real-time acquisition of real-time position information of each heat source area in the fence area, so as to timely adjust the position of the first heat source area and the second heat source area. When the first heat source area changes, the second heat source area must also change with the first heat source area. For example, if the infant is in area A in the fence area at the first moment, then area A is recorded as the first heat source area, and the surrounding area of ​​A is recorded as the second heat source area. At the second moment, the infant moves from area A to area B, and area B is part of the surrounding area of ​​A. At this time, area B is the first heat source area, and the surrounding area of ​​B is the second heat source area. Since the initial temperatures of the first heat source area and the second heat source area corresponding to the second moment and the first moment are different, and the highest temperature that can be reached in the end is not necessarily the same; therefore, corresponding temperature change thresholds are configured for each heat source area at different moments; when the temperature change of any heat source area exceeds the temperature change threshold, the change is The sleep video corresponding to the time period is analyzed to determine whether the infant's sleeping position has changed and whether there is an abnormal sleeping position. Although the above patent also combines visible light images and thermal infrared images to determine the infant's sleeping position, and discloses a technical solution of first determining whether to perform image analysis by the temperature change of the heat source area, and then determining the infant's sleeping position by image analysis, however, video image analysis has certain limitations. For example, factors such as insufficient light and frequent movements of infants and young children will affect the effect of image analysis, making the recognition of side sleeping position inaccurate; at the same time, abnormal side sleeping position may not lead to obvious changes in visible light image features, such as slight head tilt or change in the side of the body. These subtle changes are difficult to be accurately captured and recognized by the video image analysis system, so that the side sleeping position problem of infants and young children cannot be discovered in time; in summary, although the above patent uses advanced technical means to recognize the sleeping position of infants and young children, there are still many challenges and limitations in practical applications, which need to be further improved and optimized to improve the accuracy and comprehensiveness of side sleeping position recognition.

[0005] Therefore, how to accurately identify the side sleeping position of infants and young children is an urgent problem to be solved. Summary of the invention

[0006] In view of this, the present invention provides a method, device and equipment for identifying the side sleeping position of infants based on thermal infrared data analysis, so as to solve the problem that the prior art cannot accurately identify the side sleeping position of infants.

[0007] The technical solution adopted by the present invention is:

[0008] In a first aspect, the present invention provides a method for identifying side sleeping of infants based on thermal infrared data analysis, the method comprising:

[0009] Obtain the temperature value of the pixel point in the infant head area in the thermal infrared real-time image in the infant care scene;

[0010] Determining real-time statistical characteristic information corresponding to the heat distribution in the head area of ​​the infant based on the temperature value;

[0011] Calculating the real-time statistical characteristic information to obtain a skew coefficient of heat distribution;

[0012] The side sleeping position of an infant or young child is identified based on the skew coefficient and preset first, second and third thresholds corresponding to the side sleeping position, wherein the side sleeping position includes light side sleeping, moderate side sleeping or heavy side sleeping, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold.

[0013] Preferably, before acquiring the temperature value of the pixel point in the infant head area in the thermal infrared real-time image in the infant care scene, the method further includes:

[0014] Acquire thermal infrared real-time images and visible light real-time images in infant care scenarios;

[0015] Performing target detection on the visible light real-time image, and outputting first position information of the infant's head;

[0016] Determine second position information of the infant's head in the thermal infrared real-time image based on a preset mapping relationship between the thermal infrared real-time image and the visible light real-time image in combination with the first position information;

[0017] The temperature values ​​of the pixels in the infant's head area are obtained according to the second position information.

[0018] Preferably, performing target detection on the visible light real-time image and outputting first position information of the infant's head includes:

[0019] Inputting the real-time visible light image into a pre-trained target detection model, and outputting initial position information of the infant's head;

[0020] According to the initial position information of the infant's head, a target tracking algorithm is used to track the infant's head in real time and output the first position information.

[0021] Preferably, before determining the second position information of the infant's head in the thermal infrared real-time image based on the preset mapping relationship between the thermal infrared real-time image and the visible light real-time image in combination with the first position information, the method includes:

[0022] Acquire thermal infrared sample images and visible light sample images in infant care scenarios;

[0023] Performing threshold segmentation and corner point detection on the visible light sample image to obtain first vertex position information of a preset reference object;

[0024] Performing threshold segmentation and straight line detection on the thermal infrared sample image to obtain second vertex position information of a preset reference object;

[0025] The mapping relationship is determined according to the first vertex position information and the second vertex position information.

[0026] Preferably, after determining the real-time statistical characteristic information corresponding to the heat distribution in the infant's head region according to the temperature value, the method further includes:

[0027] Calculating the real-time statistical characteristic information to obtain a uniformity index of heat distribution;

[0028] If the uniformity index is less than a preset uniformity threshold corresponding to a normal sleeping position, the normal sleeping position of the infant is identified;

[0029] Obtaining statistical feature template information corresponding to the heat distribution in the head area of ​​the infant in the frontal sleeping position and the area of ​​the infant's face area;

[0030] The real-time head area of ​​the infant is obtained, and the prone sleeping position of the infant is identified based on the statistical feature template information, the face area of ​​the infant, the head area of ​​the infant and the real-time statistical feature information.

[0031] Preferably, the identifying the side sleeping position of an infant or young child according to the deflection coefficient and a preset first threshold, a second threshold and a third threshold corresponding to the side sleeping position comprises:

[0032] If the skew coefficient is greater than the first threshold value and less than the second threshold value, the infant is identified as having a mild side sleep;

[0033] If the skew coefficient is greater than or equal to the second threshold and less than the third threshold, the infant is identified as having moderate side sleeping;

[0034] If the skew coefficient is greater than or equal to the third threshold, the infant is identified as having severe side sleeping.

[0035] Preferably, the acquiring of the real-time head area of ​​the infant and identifying the prone sleeping position of the infant based on the statistical feature template information, the face area of ​​the infant, the head area of ​​the infant and the real-time statistical feature information comprises:

[0036] Determining a first evaluation value according to the area of ​​the infant's head region and the area of ​​the infant's face region;

[0037] Determine a second evaluation value according to the real-time mean value in the real-time statistical feature information and the target mean value in the statistical feature template information;

[0038] The real-time standard deviation in the real-time statistical feature information is obtained, and the prone sleeping position of the infant is identified based on the first evaluation value, the second evaluation value and the real-time standard deviation.

[0039] In a second aspect, the present invention provides a device for identifying side sleeping of infants based on thermal infrared data analysis, the device comprising:

[0040] A temperature value acquisition module, used to obtain the temperature value of the pixel point in the infant head area in the thermal infrared real-time image in the infant care scene;

[0041] A real-time statistical characteristic information extraction module, used to determine the real-time statistical characteristic information corresponding to the heat distribution in the head area of ​​the infant according to the temperature value;

[0042] A skew coefficient calculation module, used to calculate the real-time statistical characteristic information to obtain the skew coefficient of the heat distribution;

[0043] A side sleeping classification module is used to identify the side sleeping posture of infants and young children based on the skew coefficient and a preset first threshold, a second threshold and a third threshold corresponding to the side sleeping posture, wherein the side sleeping posture includes light side sleeping, moderate side sleeping or heavy side sleeping, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold. In a third aspect, an embodiment of the present invention further provides an electronic device, comprising: at least one processor, at least one memory and computer program instructions stored in the memory, and when the computer program instructions are executed by the processor, the method of the first aspect in the above-mentioned embodiment is implemented.

[0044] In a fourth aspect, an embodiment of the present invention further provides a storage medium on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the method of the first aspect in the above-mentioned embodiment is implemented.

[0045] In summary, the beneficial effects of the present invention are as follows:

[0046] The present invention provides a method, device and equipment for identifying side sleeping of infants based on thermal infrared data analysis. The method includes: obtaining the temperature value of pixel points in the infant's head area in the thermal infrared real-time image in the infant care scene; determining the real-time statistical feature information corresponding to the heat distribution in the infant's head area based on the temperature value; calculating the real-time statistical feature information to obtain the skewness coefficient of the heat distribution; identifying the side sleeping position of the infant based on the skewness coefficient and a preset first threshold, a second threshold and a third threshold corresponding to the side sleeping position, wherein the side sleeping position includes mild side sleeping, moderate side sleeping or severe side sleeping, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold. The present invention obtains thermal infrared real-time images in infant care scenarios, and identifies the side sleeping position of infants based on the temperature value and heat distribution deviation coefficient in the head area, which has significant advantages. First, by real-time monitoring of the heat distribution of the infant's head through thermal infrared technology, the temperature change and the degree of deviation when the infant is sleeping on his side can be accurately captured. Secondly, multiple thresholds (first, second, and third thresholds) are set, which can refine the side sleeping position into mild, moderate, and severe side sleeping, providing more precise posture judgment. This multi-level threshold judgment method not only improves the accuracy of recognition, but also can distinguish different degrees of side sleeping postures, which helps caregivers to take more targeted measures and reduce the risks caused by improper side sleeping postures for infants. Compared with the prior art, the solution effectively solves the problem of inability to accurately identify side sleeping postures, thereby greatly improving the reliability and safety of infant sleep monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the drawings required for use in the embodiment of the present invention will be briefly introduced below. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work, and these are all within the protection scope of the present invention.

[0048] Figure 1 This is a schematic diagram of the overall working process of the method for identifying side sleeping of infants based on thermal infrared data analysis in Example 1 of the present invention;

[0049] Figure 2 This is a schematic diagram of a process of performing target detection on the visible light real-time image in Embodiment 1 of the present invention;

[0050] Figure 3 This is a schematic diagram of a process for obtaining a mapping relationship in Embodiment 1 of the present invention;

[0051] Figure 4 This is a schematic diagram of a process of analyzing thermal infrared data in the head area of ​​an infant in a thermal infrared real-time image to identify the real-time sleeping position of the infant in Example 1 of the present invention;

[0052] Figure 5 This is a schematic diagram of a process for analyzing the real-time statistical feature information in Embodiment 1 of the present invention;

[0053] Figure 6 This is a schematic diagram of a process for identifying the side sleeping position of an infant in Embodiment 1 of the present invention;

[0054] Figure 7 This is a schematic diagram of a process for identifying the prone sleeping position of an infant in Embodiment 1 of the present invention;

[0055] Figure 8 This is a structural block diagram of a device for identifying side sleeping of infants based on thermal infrared data analysis in Example 3 of the present invention;

[0056] Fig. 9 It is a schematic diagram of the structure of an electronic device in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution in the embodiment of the present invention will be clearly and completely described in conjunction with the drawings in the embodiment of the present invention. It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. In the description of the present invention, it should be understood that the orientation or position relationship indicated by the terms "center", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc. is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. Moreover, the term "include", "comprise" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such a process, method, article or device. In the absence of further restrictions, the elements defined by the phrase "comprising..." do not exclude the existence of other identical elements in the process, method, article or device comprising the elements. If there is no conflict, the embodiments of the present invention and the various features in the embodiments can be combined with each other, all within the protection scope of the present invention.

[0058] Example 1

[0059] See also Figure 1Embodiment 1 of the present invention discloses a method for identifying side sleeping of infants based on thermal infrared data analysis, the method comprising:

[0060] Acquire thermal infrared real-time images and visible light real-time images in infant care scenarios;

[0061] Specifically, in the infant care scenario, the thermal infrared real-time image collected by the thermal infrared camera and the visible light real-time image collected by the color camera are obtained. The thermal infrared real-time image can provide real-time monitoring of the infant's body temperature distribution. Through the image collected by the thermal infrared camera, the temperature distribution of the infant's body surface can be clearly observed, so as to timely discover abnormal conditions, such as excessive or low body temperature, which is helpful for preventing and treating diseases; secondly, the visible light real-time image provides more intuitive visual monitoring. The image collected by the 1080P color camera shows the real-time dynamics of the infant in high definition, including sleep status and behavioral activities; at night, the grayscale image collected by the 940nm infrared fill light can provide a clear monitoring picture in a low-light environment, ensuring that problems can be discovered and handled in time even in the dark. Comprehensively considering the information of the thermal infrared real-time image and the visible light real-time image, the status and behavior of the infant can be fully understood, and necessary measures can be taken in time to ensure its safety and health; at the same time, the thermal infrared camera is installed directly below the visible light camera, which can achieve spatial alignment of the two images, making it convenient for users to compare and analyze during the monitoring process, and improve the monitoring effect and efficiency.

[0062] Performing target detection on the visible light real-time image, and outputting first position information of the infant's head;

[0063] Specifically, target detection of visible light real-time images refers to the use of computer vision technology to process and analyze the images, identify the head of an infant in the visible light real-time image, and obtain the corresponding position information as the first position information of the infant's head. In the infant care scenario, it is first necessary to use an appropriate target detection algorithm, including a convolutional neural network (CNN) and YoloV8s, to detect the infant's head in the image. After the head is detected, the position coordinates of the head in the image are calculated to output the first position information of the infant's head. The first position information of the infant's head is usually given in the form of pixel coordinates or normalized coordinates, which are used to indicate the relative position of the head in the image, thereby helping guardians to observe and locate the position of the infant in real time to ensure their safety.

[0064] In one embodiment, see Figure 2 , the performing target detection on the visible light real-time image and outputting the first position information of the infant's head comprises:

[0065] Inputting the real-time visible light image into a pre-trained target detection model, and outputting initial position information of the infant's head;

[0066] Specifically, first, the real-time visible light image is input into the pre-trained target detection model, and the YoloV8s model is used as the target detector, wherein YoloV8s is an efficient target detection model that can realize real-time processing while maintaining a high detection accuracy; the real-time visible light image is input into the YoloV8s model, and the target detection model will process the image and detect the infant head target therein, and the detection results output by the YoloV8s model include the position coordinates and category information of the target; the detection results classified as infant head are filtered out, and their position information, that is, the initial position information of the infant head, is extracted.

[0067] According to the initial position information of the infant's head, a target tracking algorithm is used to track the infant's head in real time, and the first position information of the infant's head is output.

[0068] Specifically, the DeepSort algorithm is used to track the infant's head, wherein DeepSort is a target tracking algorithm based on deep learning, which can accurately track the target in real time. The initial position information of the infant's head output by the YoloV8s model is used as the starting state of the tracking target. In the subsequent video frames, the DeepSort algorithm is used to track the infant's head target in each frame, and the state of the target is updated by matching the target features and motion information, and its position in the next frame is predicted; the result output by the DeepSort algorithm is the real-time tracking information of the infant's head, including its position, speed and other state information. Through the above process, the detection and tracking of the infant's head in the real-time visible light image is realized, the first position information of the infant's head is obtained, and the real-time monitoring and positioning of the infant's behavior is realized, which provides effective technical support for ensuring the safety of infants.

[0069] Determine second position information of the infant's head in the thermal infrared real-time image based on a preset mapping relationship between the thermal infrared real-time image and the visible light real-time image in combination with the first position information;

[0070] Specifically, according to the preset mapping relationship, the first position information of the infant's head obtained in the visible light real-time image is converted into the corresponding thermal infrared real-time image through the mapping relationship. The mapping relationship describes the spatial relationship between the visible light image and the thermal infrared image to ensure that the same target is located in the images under two different spectra. The output second position information of the infant's head represents the head position corresponding to the first position information in the thermal infrared image, which can realize the positioning and tracking of the infant's head in the thermal infrared image, further improving the comprehensiveness and accuracy of the monitoring system.

[0071] In one embodiment, see Figure 3 , before determining the second position information of the infant's head in the thermal infrared real-time image based on the preset mapping relationship between the thermal infrared real-time image and the visible light real-time image in combination with the first position information, the method includes:

[0072] Acquire thermal infrared sample images and visible light sample images in infant care scenarios;

[0073] Specifically, two identical square heat source black markers are placed in the fields of view of the thermal infrared camera and the color camera as preset reference objects to ensure that they can be clearly captured in both spectral images, and then the thermal infrared sample images and visible light sample images collected by the two cameras in the infant care scenario are obtained.

[0074] Performing threshold segmentation and corner point detection on the visible light sample image to obtain first vertex position information of a preset reference object;

[0075] Specifically, firstly, the outline of the square marker is extracted from the visible light image by means of the threshold segmentation technology, so that the edge of the marker can be clearly detected; wherein, the threshold segmentation technology is a commonly used image segmentation method, which is used to segment different areas in an image into parts with specific features or characteristics, and its basic idea is to compare the pixel grayscale value in the visible light sample image with one or more preset grayscale thresholds, and divide the pixels into different categories or areas according to the comparison results; then, the Harris corner detection algorithm is used to find corner points in the extracted area outline, and these corner points correspond to the vertex positions of the two square markers. The 8 vertices obtained by the corner detection algorithm, namely, Point1 (p1, p2, ... p8), are used to determine the first vertex position information of the preset reference object.

[0076] Performing threshold segmentation and straight line detection on the thermal infrared sample image to obtain second vertex position information of a preset reference object;

[0077] Specifically, in the thermal infrared sample image, the high-temperature area segmentation method is used to find the area representing the heat source, that is, the position of the two square markers. Then, the line segments of the two square markers are extracted through the Hough line detection algorithm. Because the square markers in the thermal infrared sample image will show different heat distributions, the eight intersection points Poi nt2 (p1, p2, ... p8) of the two square markers are determined by the intersection of the detected line segments, thereby determining the second vertex position information of the preset reference object.

[0078] The mapping relationship is determined according to the first vertex position information and the second vertex position information.

[0079] Specifically, the visible light sample image provides image information under the visible spectrum, and the first vertex position information of the square marker is obtained by using the contour and corner detection method; while the thermal infrared sample image provides heat distribution information, and the intersection position of the square marker is obtained as the second vertex position information through high temperature area segmentation and straight line detection. Set the affine transformation matrix of the thermal infrared camera and the visible light camera to F, and get the following equation: Point2 = F*Point1, the affine transformation matrix F has a total of 4 parameters (rotation angle, scaling factor, and 2 translation parameters); the least squares method is used to solve the above equation to get the affine transformation matrix F. Through the above steps, the markers in the visible light image and the markers in the thermal infrared image are associated, and the spatial mapping relationship between the two is established, so that the same target can be located in images under different spectra.

[0080] Based on the second position information, the thermal infrared data in the infant's head area in the thermal infrared real-time image is analyzed to identify the infant's real-time sleeping position.

[0081] Specifically, from the real-time thermal infrared image, the position and size of the infant's head area are determined according to the second position information of the infant's head, and then the thermal infrared data in the head area is extracted, and feature information related to the infant's sleeping position is extracted from the thermal infrared data, wherein the feature information includes at least temperature distribution features and shape features of the head area. Different sleeping positions are classified using the extracted feature information to identify the infant's real-time sleeping position, wherein the real-time sleeping position includes at least sleeping on the front, sleeping on the side, and sleeping on the stomach.

[0082] In one embodiment, see Figure 4 , analyzing the thermal infrared data in the head area of ​​the infant in the thermal infrared real-time image based on the second position information to identify the real-time sleeping position of the infant includes:

[0083] Acquiring temperature values ​​of pixel points in the infant's head area according to the second position information;

[0084] Specifically, according to the second position information of the infant's head, the infant's head area is located in the thermal infrared real-time image, and the temperature value of each pixel in the area is obtained. The temperature value of each pixel in the head area is expressed as T(x, y), where (x, y) represents the position of the pixel in the image coordinate system, that is, the row and column of the pixel. The thermal value T(x, y) defined in this way can be used to describe the temperature distribution at different positions in the head area, providing a basis for further analysis and application.

[0085] Determining real-time statistical characteristic information corresponding to the heat distribution in the head area of ​​the infant based on the temperature value;

[0086] Specifically, based on the temperature value, real-time statistical characteristic information is determined for the heat distribution in the head area of ​​the infant, such as the average temperature, the highest temperature, the lowest temperature, and the standard deviation of the temperature distribution. The real-time statistical characteristic information can help analyze the overall heat situation in the head area of ​​the infant, understand the temperature change trend and range in the head area, and provide a basis for health monitoring and abnormality detection. It also provides an important reference for caregivers so that they can take necessary measures in a timely manner.

[0087] The real-time statistical characteristic information is analyzed to identify the real-time sleeping position of the infant, wherein the real-time sleeping position at least includes sleeping straight, sleeping on the side, and sleeping on the stomach.

[0088] Specifically, by analyzing real-time statistical feature information, the real-time sleeping position of infants and young children can be identified, including common sleeping positions such as sleeping on the front, sleeping on the side, and sleeping on the prone. That is to say, by monitoring the heat distribution in the head area and combining the characteristic temperature distribution of the head area of ​​infants and young children in different sleeping positions, for example, the temperature distribution of the head is more uniform when sleeping on the front, the temperature distribution of the head may be biased to one side when sleeping on the side, and a higher local temperature may be shown when sleeping on the prone. Using these characteristics, the real-time sleeping position of infants and young children can be identified, providing guardians with timely sleep status monitoring and health management.

[0089] In one embodiment, see Figure 5 , the real-time statistical feature information is analyzed to identify the real-time sleeping position of the infant, wherein the real-time sleeping position at least includes front sleeping, side sleeping and prone sleeping, including:

[0090] Calculating the real-time statistical characteristic information to obtain a uniformity index and a skew coefficient of heat distribution;

[0091] Specifically, the real-time statistical feature information is calculated. First, the heat value of each pixel in the head area is calculated, and the mean value μ of these values ​​is obtained. T and standard deviation σ T , and the calculation formulas are:

[0092]

[0093] Where N is the total number of pixels in the head area, T i is the temperature value of the i-th pixel; the mean μ T reflects the average temperature of the head area as a whole, while the standard deviation σ T It reflects the degree of dispersion of heat distribution, that is, the degree of change of temperature value. Then, by T With mean μ T Divide them to get the uniformity index U T This index is used to describe the uniformity of temperature distribution in the head area. T The smaller it is, the more uniform the heat distribution is, that is, the temperature change in the head area is not large, which is closer to the normal sleeping posture. Next, calculate the deflection coefficient S T , the formula is:

[0094]

[0095] If the uniformity index is less than a preset uniformity threshold corresponding to a normal sleeping position, the normal sleeping position of the infant is identified;

[0096] Specifically, the uniformity threshold corresponding to the normal sleeping posture is obtained as 0.2. T When it is less than 0.2, the head is considered to be in a normal sleeping posture. This threshold is set based on the analysis and consideration of the general characteristics of the head temperature distribution in a normal sleeping posture. By statistically analyzing the heat distribution in the head area, the statistical characteristics of the mean and standard deviation are extracted, and the uniformity index U is defined. T As a quantitative indicator describing the uniformity of temperature distribution, real-time identification of the sleeping position of the infant's head can be achieved by comparing it with the preset normal sleeping position threshold. This method based on the statistical characteristics of temperature distribution can provide a simple and effective means for infant sleep monitoring and provide guardians with real-time monitoring and evaluation of sleep status.

[0097] Identifying the side sleeping position of an infant or young child according to the skew coefficient and a preset first threshold, a second threshold, and a third threshold corresponding to the side sleeping position, wherein the side sleeping position includes light side sleeping, moderate side sleeping, or heavy side sleeping, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold;

[0098] Specifically, according to the skew coefficient and the preset second, third and third thresholds corresponding to the side sleeping posture, the side sleeping posture of the infant is identified, and the side sleeping posture is subdivided into mild side sleeping, moderate side sleeping or severe side sleeping, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold. By comparing and judging the skew coefficient, the side sleeping posture of the infant can be divided into different degrees, and detailed information about the infant's sleeping posture preference and sleep quality can be provided to the guardian accordingly to support more detailed monitoring and attention. By further subdividing the side sleeping posture, more detailed and accurate sleep information can be provided to help the guardian better understand the infant's sleeping habits and posture preferences. By classifying side sleeping positions into mild, moderate and severe, the sleeping position of infants and young children can be assessed more accurately, potential discomfort or poor posture can be identified, and timely adjustment measures can be taken to help prevent insufficient sleep and poor sleep quality. In particular, considering that different degrees of side sleeping positions may cause infants and young children to sleep on their stomachs, for example, severe side sleeping is more likely to cause sleeping on their stomachs than mild side sleeping, and sleeping on their stomachs may increase the risk of suffocation. Subdividing side sleeping positions can help guardians more accurately assess the risk of suffocation and take necessary preventive measures, thereby improving the safety and health of infants and young children.

[0099] In one embodiment, see Figure 6 The step of identifying the side sleeping position of an infant or young child based on the deflection coefficient and a preset first threshold, a second threshold, and a third threshold corresponding to the side sleeping position includes:

[0100] If the skew coefficient is greater than the first threshold value and less than the second threshold value, the infant is identified as having a mild side sleep;

[0101] Specifically, if the skew coefficient is greater than the first threshold and less than the second threshold, for example, the first threshold is set to 0.1 and the second threshold is set to 0.5, the infant can be identified as a mild side sleeper. This indicates that the infant has a certain degree of side sleeping tendency during sleep, but the posture may not be very obvious, which is a mild side sleeping situation.

[0102] If the skew coefficient is greater than or equal to the second threshold and less than the third threshold, the infant is identified as having moderate side sleeping;

[0103] Specifically, if the skew coefficient is greater than or equal to the second threshold and less than the third threshold, for example, the third threshold is set to 1.0, the infant can be identified as a moderate side sleeper. This means that the infant presents an obvious side-lying posture during sleep, but it has not reached an extreme degree. Moderate side sleeping may bring a certain sense of comfort to infants, but it is also necessary to observe whether their sleeping state is normal, so it is necessary to remind the user to ensure the comfort and support of their head and neck.

[0104] If the skew coefficient is greater than or equal to the third threshold, the infant is identified as having severe side sleeping.

[0105] Specifically, if the skew coefficient is greater than or equal to the third threshold, the infant can be identified as a severe side sleeper. This indicates that the infant's side sleeping posture is very obvious and has a strong tendency. Severe side sleeping may cause a certain degree of pressure on the infant's body, especially the head and neck. Therefore, it is necessary to remind the user to pay special attention to whether their sleeping posture will cause discomfort or adverse consequences, and adjust their sleeping posture in time to ensure their health and safety.

[0106] Obtaining statistical feature template information corresponding to the heat distribution in the head area of ​​the infant in the frontal sleeping position and the area of ​​the infant's face area;

[0107] Specifically, first, in the face-up sleeping position, the statistical feature template information corresponding to the heat distribution in the head area of ​​the infant is analyzed to obtain the average temperature value μ of the area. Tface , and calculate the area S1 of the head region. In the front face state, the head region is usually the part directly opposite to the infant's face, that is, S1 corresponds to the infant's face region.

[0108] The real-time head area of ​​the infant is obtained, and the prone sleeping position of the infant is identified based on the statistical feature template information, the face area of ​​the infant, the head area of ​​the infant and the real-time statistical feature information.

[0109] Specifically, first, the real-time area data of the infant's head area is obtained. Secondly, the obtained head area data is combined with the pre-set statistical feature template information and the real-time statistical feature information of the infant's face area and head area to conduct a comprehensive analysis to ensure accurate recognition of the infant's posture. The ultimate goal is to identify whether the infant is in a prone sleeping position, which is crucial for guardians because the prone sleeping position may increase the risk of suffocation and breathing difficulties. Therefore, this real-time monitoring and identification method can help guardians take timely measures to ensure the health and safety of infants.

[0110] In one embodiment, see Figure 7 The acquiring of the real-time head area of ​​the infant and the identification of the prone sleeping position of the infant based on the statistical feature template information, the face area of ​​the infant, the head area of ​​the infant and the real-time statistical feature information comprises:

[0111] Determining a first evaluation value according to the area of ​​the infant's head region and the area of ​​the infant's face region;

[0112] Determine a second evaluation value according to the real-time mean value in the real-time statistical feature information and the target mean value in the statistical feature template information;

[0113] The real-time standard deviation in the real-time statistical feature information is obtained, and the prone sleeping position of the infant is identified based on the first evaluation value, the second evaluation value and the real-time standard deviation.

[0114] Specifically, when the baby is sleeping completely on his stomach, the body temperature of the back of the head is significantly lower than that of the face, and the temperature difference of the back of the head is also small. The area of ​​the back of the head and the face are similar. In the real-time sleeping state of the infant, the temperature distribution of the head area is obtained through the thermal infrared real-time image, and the average temperature value μ of the head area in the current state is calculated. Tback and standard deviation σ back , and obtain the area S2 of the infant's head area; finally, the prone sleeping state is judged according to the following conditions: the difference between the head area area in the current state and the head area area in the front face state is judged to determine the first evaluation value. If the difference is less than 20% of the head area in the front face state, that is, |S2-S1| / S1<0.2, it is considered that the head position remains relatively stable and may be in the prone sleeping state; then, the change of the head area temperature in the current state is judged. If the standard deviation σ back <1, indicating that the temperature distribution is relatively stable, and the patient may be in a prone sleeping state; finally, calculate the average temperature value μ in the face-up state. Tface The average temperature value μ in the current state Tback The difference ΔT between them is used to determine the second evaluation value. If ΔT is greater than 20% of the average temperature value in the front face state, that is, 0.2*μ Tface ), it is considered that the head temperature is significantly reduced and the infant may be in a prone sleeping state; if the above three conditions are met at the same time, it can be considered that the infant is in a prone sleeping state. Prone sleeping is an unhealthy posture for infants and young children to sleep, which may increase the risk of suffocation and dyspnea. Through this solution, it is possible to timely detect that the infant is in a prone sleeping state, which helps the guardian to take timely measures to avoid adverse consequences. By comprehensively considering multiple factors such as the area of ​​the head area, the stability of the temperature distribution, and the temperature change, the judgment of the prone sleeping state is more comprehensive and accurate, thereby reducing the situation of misidentification and missed identification and improving the accuracy of identification; at the same time, the above prone sleeping identification scheme is based on non-invasive monitoring methods such as thermal infrared images, and does not require direct contact with infants and young children, avoiding interference with their sleep, which is very important for maintaining the sleeping state and comfort of infants and young children. By real-time monitoring of the prone sleeping state, guardians can better understand the sleeping habits and posture preferences of infants and young children, which helps to formulate personalized sleep management plans. For example, the sleeping posture or sleeping environment of infants and young children is adjusted according to the identification results to improve sleep quality and safety.

[0115] Example 2

[0116] On the basis of Example 1, the real-time sleeping position of the care object has been identified, wherein the real-time sleeping position includes: straight sleeping, light side sleeping, moderate side sleeping, heavy side sleeping and prone sleeping. In order to improve the sleep quality, this Example 2 develops a sleeping position correction plan by monitoring different sleeping positions and durations. Based on the second position information, the thermal infrared data in the head area of ​​the infant in the thermal infrared real-time image is analyzed. After identifying the real-time sleeping position of the infant, it also includes:

[0117] According to the total sleep time of the infant and the target sleep time corresponding to each real-time sleeping position, determine the proportion of each real-time sleeping position to the total sleep time;

[0118] Specifically, the start and end timestamps of various real-time sleeping positions are obtained, and the target sleep duration corresponding to each real-time sleeping position is calculated through the start and end timestamps. Then, based on the total sleep duration of the infant, the proportion of time occupied by each sleeping position in the entire sleep cycle is determined. This calculation result helps guardians and medical experts understand the sleep patterns of infants and assess whether there are potential risks.

[0119] For example, if an infant is detected sleeping on his / her stomach, the system will immediately send out an alarm through a connected device (such as a smartphone or other mobile device) to prompt the guardian to adjust the infant's sleeping position, as sleeping on the stomach is considered to increase the risk of sudden infant death syndrome. In addition, if a certain sleeping position, such as severe side sleeping, accounts for an abnormally high proportion, it may have a negative impact on the spinal health of infants and young children, and adjustments will also be recommended. This real-time feedback mechanism not only enhances the safety of infants' sleep, but also provides parents with valuable guidance to help them take correct care measures to ensure the overall health and development of infants and young children.

[0120] Various safety reminders are issued based on the proportion of each real-time sleeping posture to the total sleep time and the preset proportion threshold.

[0121] In one embodiment, the issuing of multiple safety reminders based on the ratio of each real-time sleeping posture to the total sleep time and a preset ratio threshold includes:

[0122] If the proportion of the upright sleeping posture to the total sleep time is greater than the preset first ratio threshold, or the proportion of the light side sleeping posture to the total sleep time is greater than the preset first ratio threshold, no safety reminder will be issued;

[0123] Specifically, when the proportion of the infant's upright sleeping or mild side sleeping posture in the total sleep time reaches a safe and healthy level, such as 50% each, this is generally regarded as a healthy sleeping posture distribution. If these proportions are higher than the preset first proportion threshold, such as more than 60% each, the infant's sleeping posture is considered normal and no safety reminder is required. This indicates that the infant has experienced enough upright sleeping and healthy side sleeping during sleep, which is conducive to promoting balanced physical development and reducing the risk of SIDS.

[0124] If the proportion of the moderate side sleeping posture to the total sleep time is greater than a preset second proportion threshold, a first safety reminder is issued;

[0125] Specifically, if the proportion of moderate side sleeping exceeds the preset second proportion threshold, such as exceeding 30%, the system will issue a first safety reminder. Moderate side sleeping will cause one side of the body to bear greater pressure and affect the health of the spine. In this case, the first safety reminder will suggest that the guardian gently adjust the posture of the infant to avoid prolonged unilateral pressure.

[0126] If the proportion of severe side sleeping posture to the total sleep time is greater than the preset third ratio threshold, a second safety reminder is issued;

[0127] Specifically, when the proportion of severe side sleeping posture exceeds the third ratio threshold, such as exceeding 20%, a second safety reminder will be issued. Severe side sleeping will cause the infant's body to bend excessively and increase the risk of scoliosis, which requires timely intervention. The second safety reminder will prompt the guardian to adjust the infant's posture and will also recommend the use of specific support materials (such as pillows) to help maintain a safer sleeping position.

[0128] If the real-time sleeping position of the infant is identified as prone sleeping position, a third safety reminder is issued; wherein the first ratio threshold is greater than the second ratio threshold, and the second ratio threshold is greater than the third ratio threshold.

[0129] Specifically, prone sleeping is a posture that requires special attention because it is associated with an increased risk of sudden infant death syndrome. Once an infant is detected in a prone sleeping position, regardless of how long it lasts, a third safety reminder will be issued immediately. This reminder is urgent and aims to allow the guardian to take quick action to adjust the infant's sleeping position to ensure that his or her breathing is unobstructed and safe.

[0130] Example 3

[0131] See also Figure 8 Embodiment 3 of the present invention further provides a device for identifying side sleeping of infants based on thermal infrared data analysis, the device comprising:

[0132] An image acquisition module, used to acquire real-time thermal infrared images and real-time visible light images in infant care scenarios;

[0133] A target detection module, used to perform target detection on the real-time visible light image and output first position information of the infant's head;

[0134] A mapping module, configured to determine second position information of the infant's head in the thermal infrared real-time image based on a preset mapping relationship between the thermal infrared real-time image and the visible light real-time image in combination with the first position information;

[0135] The sleeping posture recognition module is used to analyze the thermal infrared data in the head area of ​​the infant in the thermal infrared real-time image according to the second position information, and recognize the real-time sleeping posture of the infant.

[0136] Specifically, a device for identifying side sleeping of infants based on thermal infrared data analysis provided in an embodiment of the present invention is used, and the device includes: an image acquisition module, used to acquire thermal infrared real-time images and visible light real-time images in infant care scenarios; a target detection module, used to perform target detection on the visible light real-time image, and output first position information of the infant's head; a mapping module, used to determine the second position information of the infant's head in the thermal infrared real-time image based on a preset mapping relationship between the thermal infrared real-time image and the visible light real-time image, combined with the first position information; a sleeping position recognition module, used to analyze the thermal infrared data in the infant's head area in the thermal infrared real-time image based on the second position information, and identify the infant's real-time sleeping position. This device combines two different information sources, thermal infrared real-time images and visible light real-time images. The visible light image provides accurate detection of the infant's head position, while the thermal infrared image provides thermal infrared data in the infant's head area. By fusing these two types of information, the sleeping position can be analyzed more comprehensively; the visible light image is used for target detection to accurately locate the infant's head position, providing an accurate reference point for subsequent thermal infrared image analysis; the infant's head position information and thermal infrared data in the thermal infrared real-time image are used to conduct a more in-depth analysis of the infant's real-time sleeping position. The thermal infrared data can reflect the temperature distribution in the infant's head area, thereby accurately identifying different sleeping positions, such as side sleeping, supine sleeping, prone sleeping, etc. In summary, this device improves the accuracy and comprehensiveness of sleeping position recognition by fusing multiple information sources and combining precise position positioning and thermal infrared data analysis, which helps to ensure the sleep safety of infants.

[0137] Example 4

[0138] In addition, combined Figure 1 The method for identifying side sleeping of infants based on thermal infrared data analysis of the first embodiment of the present invention described above can be implemented by an electronic device. Fig. 9 A schematic diagram of the hardware structure of an electronic device provided in Embodiment 4 of the present invention is shown.

[0139] The electronic device may include a processor and a memory storing computer program instructions.

[0140] Specifically, the processor may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present invention.

[0141] The memory may include a large capacity memory for data or instructions. By way of example and not limitation, the memory may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory may be inside or outside a data processing device. In a specific embodiment, the memory is a non-volatile solid-state memory. In a specific embodiment, the memory includes a read-only memory (ROM). In appropriate cases, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM) or a flash memory or a combination of two or more of these.

[0142] The processor implements any one of the methods for identifying side sleeping of infants based on thermal infrared data analysis in the above-mentioned embodiments by reading and executing computer program instructions stored in the memory.

[0143] In one example, the electronic device may further include a communication interface and a bus. Fig. 9 As shown, the processor, memory, and communication interface are connected via a bus and communicate with each other.

[0144] The communication interface is mainly used to implement communication between the modules, devices, units and / or equipment in the embodiments of the present invention.

[0145] The bus includes hardware, software or both, and the parts of the device are coupled to each other. For example, but not limitation, the bus may include an accelerated graphics port (AGP) or other graphics bus, an enhanced industrial standard architecture (EISA) bus, a front-side bus (FSB), a hypertransport (HT) interconnection, an industrial standard architecture (ISA) bus, an infinite bandwidth interconnection, a low pin count (LPC) bus, a memory bus, a microchannel architecture (MCA) bus, a peripheral component interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a serial advanced technology attachment (SATA) bus, a video electronics standard association local (VLB) bus or other suitable bus or a combination of two or more of these. In suitable cases, the bus may include one or more buses. Although the embodiments of the present invention describe and illustrate specific buses, the present invention considers any suitable bus or interconnection.

[0146] Example 5

[0147] In addition, in combination with the infant side sleeping identification method based on thermal infrared data analysis in the above embodiment 1, the present invention can also provide a computer readable storage medium for implementation in embodiment 5. The computer readable storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any one of the infant side sleeping identification methods based on thermal infrared data analysis in the above embodiments is implemented.

[0148] In summary, the embodiments of the present invention provide a method, device and equipment for identifying side sleeping of infants based on thermal infrared data analysis.

[0149] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0150] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0151] It should also be noted that the exemplary embodiments mentioned in the present invention describe some methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiments, or in a different order from the embodiments, or several steps can be performed simultaneously.

[0152] The above is only a specific implementation of the present invention. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. It should be understood that the protection scope of the present invention is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be covered within the protection scope of the present invention.

Claims

1. A method for identifying side sleeping of infants based on thermal infrared data analysis, characterized in that: The method comprises: Obtain the temperature value of the pixel point in the infant head area in the thermal infrared real-time image in the infant care scene; Determining real-time statistical characteristic information corresponding to the heat distribution in the head region of the infant based on the temperature value; Calculating the real-time statistical characteristic information to obtain a skew coefficient of heat distribution; The side sleeping position of an infant or young child is identified based on the skew coefficient and preset first, second and third thresholds corresponding to the side sleeping position, wherein the side sleeping position includes light side sleeping, moderate side sleeping or heavy side sleeping, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold.

2. The method for identifying side sleeping of infants based on thermal infrared data analysis according to claim 1, characterized in that: Before obtaining the temperature value of the pixel point in the infant head area in the thermal infrared real-time image in the infant care scene, the method further includes: Acquire thermal infrared real-time images and visible light real-time images in infant care scenarios; Performing target detection on the visible light real-time image, and outputting first position information of the infant's head; Determine second position information of the infant's head in the thermal infrared real-time image based on a preset mapping relationship between the thermal infrared real-time image and the visible light real-time image in combination with the first position information; The temperature values ​​of the pixels in the infant's head area are obtained according to the second position information.

3. The method for identifying side sleeping of infants based on thermal infrared data analysis according to claim 2, characterized in that: The performing target detection on the visible light real-time image and outputting first position information of the infant's head comprises: Inputting the real-time visible light image into a pre-trained target detection model, and outputting initial position information of the infant's head; According to the initial position information of the infant's head, a target tracking algorithm is used to track the infant's head in real time and output the first position information.

4. The method for identifying side sleeping of infants based on thermal infrared data analysis according to claim 2, characterized in that: Before determining the second position information of the infant's head in the thermal infrared real-time image based on the preset mapping relationship between the thermal infrared real-time image and the visible light real-time image in combination with the first position information, the method includes: Acquire thermal infrared sample images and visible light sample images in infant care scenarios; Performing threshold segmentation and corner point detection on the visible light sample image to obtain first vertex position information of a preset reference object; Performing threshold segmentation and straight line detection on the thermal infrared sample image to obtain second vertex position information of a preset reference object; The mapping relationship is determined according to the first vertex position information and the second vertex position information.

5. The method for identifying side sleeping of infants based on thermal infrared data analysis according to claim 1, characterized in that: After determining the real-time statistical characteristic information corresponding to the heat distribution in the head area of ​​the infant according to the temperature value, the method further includes: Calculating the real-time statistical characteristic information to obtain a uniformity index of heat distribution; If the uniformity index is less than a preset uniformity threshold corresponding to a normal sleeping position, the normal sleeping position of the infant is identified; Obtaining statistical feature template information corresponding to the heat distribution in the head area of ​​the infant in the frontal sleeping position and the area of ​​the infant's face area; The real-time head area of ​​the infant is obtained, and the prone sleeping position of the infant is identified based on the statistical feature template information, the face area of ​​the infant, the head area of ​​the infant and the real-time statistical feature information.

6. The method for identifying side sleeping of infants based on thermal infrared data analysis according to claim 1, characterized in that: The identifying the side sleeping position of an infant or young child according to the deflection coefficient and a preset first threshold, a second threshold, and a third threshold corresponding to the side sleeping position comprises: If the skew coefficient is greater than the first threshold value and less than the second threshold value, the infant is identified as having a mild side sleep; If the skew coefficient is greater than or equal to the second threshold and less than the third threshold, the infant is identified as having moderate side sleeping; If the skew coefficient is greater than or equal to the third threshold, the infant is identified as having severe side sleeping.

7. The method for identifying side sleeping of infants based on thermal infrared data analysis according to claim 5, characterized in that: The step of obtaining the real-time area of ​​the infant's head region and identifying the infant's prone sleeping position based on the statistical feature template information, the infant's face region area, the infant's head region area and the real-time statistical feature information comprises: Determining a first evaluation value according to the area of ​​the infant's head region and the area of ​​the infant's face region; Determine a second evaluation value according to the real-time mean value in the real-time statistical feature information and the target mean value in the statistical feature template information; The real-time standard deviation in the real-time statistical feature information is obtained, and the prone sleeping position of the infant is identified based on the first evaluation value, the second evaluation value and the real-time standard deviation.

8. A device for identifying side sleeping of infants based on thermal infrared data analysis, characterized in that: The device comprises: A temperature value acquisition module, used to obtain the temperature value of the pixel point in the infant head area in the thermal infrared real-time image in the infant care scene; A real-time statistical characteristic information extraction module, used to determine the real-time statistical characteristic information corresponding to the heat distribution in the head area of ​​the infant according to the temperature value; A skew coefficient calculation module, used to calculate the real-time statistical characteristic information to obtain the skew coefficient of the heat distribution; The side sleeping classification module is used to identify the side sleeping position of infants and young children based on the skew coefficient and a preset first threshold, a second threshold and a third threshold corresponding to the side sleeping position, wherein the side sleeping position includes mild side sleeping, moderate side sleeping or severe side sleeping, wherein the first threshold is less than the second threshold, and the second threshold is less than the third threshold.

9. An electronic device, characterized in that: include: At least one processor, at least one memory and computer program instructions stored in the memory, when the computer program instructions are executed by the processor, implement the method according to any one of claims 1 to 7.

10. A storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images