An intelligent temperature detection method for infants and young children based on multi-region positioning of the head
By using multi-area head positioning technology in the infant and young children's intelligent body temperature detection system to identify and measure multiple effective temperature measurement areas of infants' heads, the accuracy of body temperature detection in various sleeping positions of infants and young children is solved, and the problem of temperature detection in the infants and young children is realized Real-time and accurate temperature monitoring.
Patent Information
- Application Number
- CN202111422493.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-11-26
- Publication Date
- 2025-02-11
- Estimated Expiration
- 2041-11-26
AI Technical Summary
The existing infant temperature detection methods are difficult to accurately measure temperature when infants and young children often change their positions, especially in various sleeping positions such as lying on their back, side and prone.
An intelligent body temperature detection system based on multi-area positioning of the head is adopted, which includes a receiving end, an intelligent analysis module, an infrared temperature measurement module, a camera module, a gimbal and a support frame. By identifying the characteristic points of the infant's head, demarcate the three temperature measurement areas of the forehead, temporal and back of the ear, and perform multi-point temperature measurement calculation and environmental compensation correction based on the infant's sleeping position.
Accurate body temperature measurement in various sleeping positions of infants and young children is achieved, meeting the real-time detection needs of infants and young children by guardians, and improving the accuracy and reliability of the detection.
Smart Images

Figure CN114216564B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of human body temperature measurement, and in particular to an intelligent body temperature detection method for infants and young children based on multi-region positioning of the head. Background Art
[0002] In recent years, it has gradually become a consensus among parents to raise children carefully. As a result, people are paying more and more attention to the physical health of infants and young children, and are willing to use technology to enhance the monitoring of their children's health. Among them, knowing the child's body temperature is the most common way to pay attention. In the past, the most common way to detect body temperature was to touch the forehead with the hand and monitor with a thermometer. Especially when the child's mental state was not good, parents would frequently check the child's body temperature.
[0003] At present, the main methods used for individual temperature measurement include mercury thermometers, electronic thermometers, ear thermometers and infrared thermometers. In recent years, non-contact, fast and accurate temperature measurement methods have become increasingly popular, especially for children. Forehead thermometers and temporal artery thermometers without safety risks and discomfort are very popular. Forehead thermometers are handheld infrared induction thermometers, which are commonly used in hospitals. They measure the temperature of the human forehead. Temporal artery thermometers use infrared scanners to measure the temperature of the temporal artery, which is as accurate as the rectal temperature of healthy individuals. The temporal artery is located on both sides of the forehead, close to the skin and passes through the forehead. It is the most practical way to measure the temperature of children under 3 years old. Even fat people can easily measure it, and it is easy to operate, especially when the child is asleep. The advantages of infrared temperature measurement are fast reading, displaying the body temperature in 1 second, without direct contact with the human body, and can also be measured when the baby is sleeping, and contact transmission is avoided. The disadvantage is that the measurement results are easily interfered by external factors, such as outdoor temperature, light, radiation, etc., so correction is required to be more accurate. In addition, the price is relatively high.
[0004] The thermometers mentioned above all need to be operated by parents, which is feasible during the day, but difficult to do at night. The current parenting concept begins to advocate that children and parents sleep in separate beds, and is gradually accepted by the younger generation of parents, so many infants and young children sleep in the same room as their parents or even have their own separate bedroom. This creates a need for continuous temperature monitoring of infants and young children, that is, it is necessary to locate the temperature measurement area of the infant and young children and measure the temperature at regular intervals. Considering that infants and young children often change their positions and postures when sleeping, there are currently designs that combine infrared temperature measurement equipment with a pan-tilt head, locate the child's face and forehead based on face recognition, and then adjust the angle of the temperature measurement equipment to perform temperature detection. However, since the temperature measurement equipment can only be fixed at a certain position next to the bed, and infants and young children often flip and roll during sleep, and their postures change, the forehead area used for temperature measurement is often not within the range of infrared temperature measurement perception. For example Figure 3The postures shown are: a for supine, b for prone, and c for side-lying. Usually, the temperature measuring device is fixed above or above the side of the infant. Only in the supine posture a can the forehead be accurately located and the temperature measured. Therefore, the current pan-tilt type temperature detection equipment measures the forehead temperature by default, without fully considering how to solve the problem of accurate temperature measurement when infants and young children often change their postures.
[0005] In addition to the forehead used for conventional non-contact temperature measurement, the temporal area on the forehead and behind the ears of infants and young children are relatively accurate temperature measurement sites. As mentioned earlier, the accuracy of temperature measurement in the temporal area is higher than that in the forehead, and behind the ear is another area worthy of attention. Measuring the temperature behind the ear can eliminate the possibility of false hypothermia caused by evaporative cooling of sweating on the skin surface, because sweating on the forehead often does not attract attention. Sweating causes the temperature of the temperature measurement area to drop, and the area directly behind the earlobe has a high blood flow required for arterial measurement, and it is the part of the body least likely to sweat, and at the same time, it is easy to measure. In the case of sweating on the forehead and complete dilation of the blood vessels in the neck, the back of the ear can also be used as the only temperature measurement point. If there is no sweating, head trauma, or it is uncertain whether there is vascular dilation in the neck, then the back of the ear cannot be used as the only temperature measurement point. If the forehead area is completely covered by a bandage, burns or lacerations, the back of the earlobe can be used to measure the temperature accurately.
[0006] In summary, therefore, it is necessary to improve and optimize the current methods for detecting infant and young child temperature. Summary of the invention
[0007] The purpose of the present invention is to overcome the above-mentioned deficiencies in the prior art and to provide an intelligent temperature detection method for infants based on multi-region positioning of the head.
[0008] The technical solution adopted by the present invention to solve the above-mentioned problem is: an intelligent body temperature detection method for infants based on multi-zone positioning of the head, which is implemented by an intelligent body temperature detection system for infants. The intelligent body temperature detection system for infants includes a receiving end, an intelligent analysis module, an infrared temperature measurement module, a camera module, a pan head and a support frame. The pan head is installed on the support frame, and the intelligent analysis module, the infrared temperature measurement module and the camera module are integrated and installed on the pan head. The camera module is used to shoot infant videos, the intelligent analysis module is used to analyze infant videos and control the infrared temperature measurement module to measure temperature according to the analysis results, and the receiving end is used to receive infant videos and temperature measurement results.
[0009] The steps of the intelligent temperature detection method for infants and young children based on multi-region positioning of the head are as follows:
[0010] Initialization: Install the infant intelligent temperature detection system above or above the infant's bed through the support frame. The installation position should ensure that the effective temperature measurement position of the infant can be identified and detected; set the infant's usual sleeping area according to the size and position of the crib, that is, the position where parents usually put their infants, and enter the coordinate range into the infant intelligent temperature detection system;
[0011] Step S1: Detect the sleeping area of the infant and confirm the position of the infant;
[0012] Step S2: Identify the effective temperature measurement area for infants and young children and perform multi-area positioning:
[0013] The specific steps are:
[0014] Step S2.1: The head posture of the infant is estimated by the Yolov3 model to determine the sleeping position of the infant. There are three sleeping positions: supine, side-lying, and prone.
[0015] Step S2.2: Based on the characteristic points of the infant's head, the forehead area A, the temporal area B and the area behind the ear C are identified and delineated. The forehead area A, the temporal area B and the area behind the ear C are the effective areas for measuring the infant's temperature;
[0016] Step S3: Perform multi-point temperature measurement calculation for the located multiple temperature measurement areas: Setting: For the supine posture, only the forehead area A can be completely identified, and the temperature can be measured through the forehead area A and the temporal area B; for the side-lying posture, only the temporal area B can be completely identified, and the temperature can be measured through the forehead area A, the temporal area B and the area C behind the ear; for the prone posture, only the area C behind the ear can be completely identified, and the temperature can be measured through the temporal area B and the area C behind the ear; It is particularly pointed out that if there are two temporal areas B in the identification area at the same time, the area B with a larger area is selected as the temporal area B for temperature measurement.
[0017] The specific steps for multi-point temperature measurement calculation in multiple temperature measurement areas are as follows:
[0018] Step S3.1: The obtained forehead area A, temporal area B and behind the ear area C are respectively measured at five points, where the five points are the center point of the area and the temperature measurement points at a position halfway from the upper, lower, left and right edges. The temperatures TA, TB and TC obtained by the temperature measurement are the average temperatures of the five points of the forehead area A, temporal area B and behind the ear area C respectively;
[0019] Step S3.2: Calculate the temperature T0 using different temperature calculation methods according to different sleeping positions:
[0020] (1) If the sleeping position is judged as supine, the temperature is determined based on TA and TB. If |TA-TB| < 0.3°C, T0 = (TA+TB) / 2; otherwise, T0 is equal to the higher temperature of TA and TB.
[0021] (2) If the sleeping position is judged as side-lying, the temperature is measured based on TA, TB, and TC. If |TA-TB| < 0.3°C, T0 = TB; otherwise, T0 is equal to the higher temperature among TA, TB, and TC.
[0022] (3) If the sleeping position is judged as prone, the temperature is measured based on TB and TC. If |TB-TC| < 0.3°C, then (TB+TC) / 2; otherwise, T0 is equal to the higher temperature of TB and TC;
[0023] Step S4: Perform environmental compensation correction on the temperature measurement result: measure the ambient temperature T1, use the compensation algorithm based on BP neural network to perform infrared temperature measurement compensation, and obtain the final temperature measurement result T;
[0024] Step S5: Process the abnormal temperature.
[0025] Preferably, in step S1, the specific method for detecting and confirming the position of the infant is:
[0026] Step S1.1: The camera module takes pictures of the area where infants and young children usually sleep, and the intelligent analysis module uses the Yolov3 model to perform human body recognition to detect whether there are infants and young children in the area;
[0027] Step S1.2: The infrared temperature measurement module detects the temperature of the identified human body to confirm whether the highest temperature is within the set temperature range, which is 34-42°C;
[0028] If both of the above conditions are met, it is determined that the infant is in the area, and the process goes to step S2;
[0029] If it is determined that the infant is not in the area, the shooting scanning range of the camera module is expanded to a range three times the length and width, and step S1 is repeated. If it is determined again that the infant is not in the expanded area, an alarm is issued.
[0030] Preferably, in step S5, the method for processing abnormal temperature is:
[0031] If the final temperature measurement result T is not within the set temperature range of 34-42℃, discard the data and re-measure. If the re-measurement result is still not within the set temperature range, the feedback temperature is wrong.
[0032] If the final temperature measurement result T is within the set temperature range of 34-42°C, but exceeds the alarm threshold, the alarm threshold is set to 37°C, an alarm is given to the guardian, and the temperature measurement time interval is reduced to half of the original.
[0033] Preferably, in step S2.2, the feature points of the infant's head based on which the setting is based include the nose tip, the corners of the eyes, the upper edge of the eyebrows, the earlobes, the hairline and the edge of the forehead, wherein except for the edge of the forehead and the hairline, the other feature points are stable feature points;
[0034] For infants and young children in the supine position, the specific method for identifying and dividing the effective temperature measurement area is as follows:
[0035] (1) Use the OpenCV library function to identify the two corners of the eyes and the nose, and calculate the central axis of the head parallel to the nose;
[0036] (2) Establish a face region coordinate system and project the face into the xy two-dimensional plane coordinate system, where the x-axis corresponds to the line connecting the two eye corners, the y-axis corresponds to the central axis of the head, and the intersection of the two axes is the origin;
[0037] (3) The two corners of the eyes are divided into horizontal coordinates x1 and x2, and lines are drawn along the y direction parallel to the central axis of the head to form the AY1 and AY2 edges;
[0038] (4) The upper edges of the eyebrows on both sides are connected to form the AX1 side;
[0039] (5) The intersection of AY1 and the top of the head contour, the intersection of AY2 and the top of the head contour, the minimum ordinate of these three points along the lowest hairline is y2, and a line is drawn along y2 parallel to the x-axis to form the AX2 edge;
[0040] (6) The area formed by the four sides of AX1, AX2, AY1, and AY2 is the frontal A area;
[0041] (7) Extend AX1 and AX2 along the x-axis to the side profile of the head, rotate the extended line downward 15 degrees, and the intersection with the side profile of the head is By1 and By2. The two adjacent vertices of area A form a quadrilateral with By1 and By2, which is the temporal area B. For the side-lying posture, the specific method for identifying and dividing the effective area for temperature measurement of infants and young children is as follows:
[0042] (1) The opencv library function can be used to identify the eye corners, nose and ear areas, and feature recognition can be used to determine the edge lines of the eye corners, eyebrows and auricles;
[0043] (2) Draw a tangent line BX1 from the point where the auricle connects to the upper side of the head to the eyebrow on the same side, draw a perpendicular line BY2 at half the distance from the tangent point, and draw a perpendicular line BY1 from the canthus of the same side perpendicular to the midline of the eye;
[0044] (3) Draw line BX2 parallel to BX1 at the intersection of BY1 and the hairline on the top of the head; if the hairline cannot be identified, draw line BX2 parallel to BX1 at the distance between the upper edge of the eyebrow and the midpoint of the top of the head;
[0045] (4) The area formed by the intersection of BX1, BX2, BY1, and BY2 is the temporal B area;
[0046] (5) On the side of the temporal area B, extend BX1 and BX2 toward the forehead, and the trapezoidal area formed by the intersection with the edge of the forehead contour is the forehead area A;
[0047] (6) The depression below the earlobe is the C area behind the ear, the height is half the size of the auricle, and the width is set to half the height. For the prone position, the specific method for identifying and dividing the effective temperature measurement area for infants and young children is as follows:
[0048] (1) The ear area can be identified using the OpenCV library function, and the corners of the eyes, eyebrow edges, auricles, and earlobes can be determined using feature recognition;
[0049] (2) The depression below the earlobe is the C zone behind the ear. Its height is half the size of the auricle and its width is half the height. The outer edge of the C zone behind the ear is an arc line.
[0050] (3) Draw a line BX1 from the point where the auricle connects to the upper side of the head to the edge of the eyebrow on the same side, and draw a perpendicular line BY2 at half the distance from the edge of the eyebrow;
[0051] (4) The area enclosed by the extension of BX1 and BY2 together with the outer edge of the head and the hairline is the temporal area B.
[0052] Compared with the prior art, the present invention has the following advantages and effects: the present invention has the following advantages and effects: the present invention has the following advantages and effects: the present invention has the following advantages and effects: the present invention has the following advantages and effects: the present invention has the following advantages and effects: the present invention has the following advantages and effects: the present invention has the following advantages and effects: the present invention has the following advantages and effects: BRIEF DESCRIPTION OF THE DRAWINGS
[0053] In order to more clearly illustrate the specific implementation methods of the present invention or the solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.
[0054] Figure 1 This is a flow chart of the intelligent temperature detection method for infants and young children based on multi-zone positioning of the head in this embodiment.
[0055] Figure 2 Schematic diagram of the structure of the intelligent temperature detection system for infants in this embodiment.
[0056] Figure 3 It shows three sleeping positions for infants and young children. In the picture, a is lying on the back, b is lying on the stomach, and c is lying on the side.
[0057] Figure 4 Indicates the effective temperature measurement area when sleeping in the supine position.
[0058] Figure 5 It illustrates the process of locating the effective temperature measurement area when sleeping in a supine position.
[0059] Figure 6 Indicates the effective temperature measurement area when sleeping on the side.
[0060] Figure 7 Illustration of the process of locating the effective temperature measurement area when sleeping in the side-lying position.
[0061] Figure 8 Indicates the effective temperature measurement area when sleeping in the prone position.
[0062] Fig. 9 It illustrates the process of locating the effective temperature measurement area when sleeping in a prone position.
[0063] Fig.10 It indicates the effective temperature measurement areas corresponding to the three sleeping positions.
[0064] Description of the accompanying drawings: receiving end 1; infrared temperature measurement module 2; camera module 3; pan / tilt head 4, support frame 5; intelligent analysis module 6. DETAILED DESCRIPTION
[0065] The present invention will be further described in detail below with reference to the accompanying drawings and by way of examples. The following examples are intended to explain the present invention but the present invention is not limited to the following examples.
[0066] Example
[0067] See also Figures 1 to 10 .
[0068] This embodiment discloses an intelligent temperature detection method for infants and young children based on multi-region positioning of the head. The detection method is implemented by an intelligent temperature detection system for infants and young children. The intelligent temperature detection system for infants and young children includes a receiving end 1, an intelligent analysis module 6, an infrared temperature measurement module 2, a camera module 3, a pan head 4 and a support frame 5. The pan head 4 is installed on the support frame 5, and the intelligent analysis module 6, the infrared temperature measurement module 2 and the camera module are integrated and installed on the pan head 4. The pan head 4 can adjust the angles of the camera lens and the temperature measurement lens. For details, refer to the prior art. The camera module 3 is used to shoot infant videos. The intelligent analysis module 6 is integrated with the intelligent algorithm involved in the detection method, which plays the role of the brain of the detection system, and is used to analyze the infant video and control the infrared temperature measurement module 2 to measure the temperature according to the analysis results. In addition, the pan head 4 is also provided with an alarm module and a wireless transmission module. The alarm information can be transmitted to the receiving end through the wireless transmission module. The receiving end 1 can be a device such as a mobile phone of a guardian, which is used to receive infant videos, temperature measurement results and alarm information.
[0069] In this embodiment, initialization is performed before temperature measurement: the infant intelligent body temperature detection system is installed above or above the side of the infant's bed through the support frame 5. The installation position must ensure that the effective temperature measurement position of the infant can be identified and detected. In order to ensure the accuracy of the temperature measurement method and the recognizability of the face, the distance from the infrared temperature measurement module 2 to the infant's head is controlled within the range of 10 cm to 100 cm; the infant's usual sleeping area is set according to the size and position of the crib, that is, the position where parents are accustomed to putting the infant, and the coordinate range is input into the infant intelligent body temperature detection system.
[0070] The steps of this intelligent temperature detection method for infants and young children based on multi-region positioning of the head are as follows:
[0071] Step S1: Detect the sleeping area of the infant and confirm the position of the infant. The specific method of detecting and confirming the position of the infant is as follows:
[0072] Step S1.1: The camera module takes pictures of the area where infants and young children usually sleep, and the intelligent analysis module uses the Yolov3 model to perform human body recognition to detect whether there are infants and young children in the area;
[0073] Step S1.2: The infrared temperature measurement module detects the temperature of the identified human body to confirm whether the highest temperature is within the set temperature range. The set temperature range is 34-42°C. This set temperature range can avoid misidentification of the toy doll;
[0074] If both of the above conditions are met, it is determined that the infant is in the area, and the process goes to step S2;
[0075] If it is determined that the infant is not in the area, the shooting scanning range of the camera module is expanded to a range three times the length and width, and step S1 is repeated. If it is determined again that the infant is not in the expanded area, an alarm is issued.
[0076] Step S2: Identify the effective temperature measurement area for infants and young children and perform multi-area positioning:
[0077] The specific steps are:
[0078] Step S2.1: The head posture of the infant is estimated by the Yolov3 model to determine the infant's sleeping position. There are three sleeping positions: supine, side-lying, and prone (see Figure 3 ), where the head posture recognition training can be based on the front and side face atlases provided by public databases such as FERET and LFW. For the pictures that have been identified as heads, if they are not the front and side faces, they are the side and back faces;
[0079] Step S2.2: Based on the feature points of the infant's head, the forehead area A, the temporal area B and the area behind the ear C are identified and delineated. The forehead area A, the temporal area B and the area behind the ear C are the effective areas for measuring the temperature of the infant. The feature points of the infant's head based on the feature points are set to include the nose tip, the corner of the eye, the upper edge of the eyebrow, the earlobe, the hairline and the edge of the forehead. Except for the edge of the forehead and the hairline, the other feature points are stable feature points.
[0080] See also Figure 4 and Figure 5 For the supine position, the specific method for identifying and dividing the effective temperature measurement area for infants and young children is as follows:
[0081] (1) Use the OpenCV library function to identify the two corners of the eyes and the nose, and calculate the central axis of the head parallel to the nose;
[0082] (2) Establish a face region coordinate system and project the face into the xy two-dimensional plane coordinate system, where the x-axis corresponds to the line connecting the two eye corners, the y-axis corresponds to the central axis of the head, and the intersection of the two axes is the origin;
[0083] (3) The two corners of the eyes are divided into horizontal coordinates x1 and x2, and lines are drawn along the y direction parallel to the central axis of the head to form the AY1 and AY2 edges;
[0084] (4) The upper edges of the eyebrows on both sides are connected to form the AX1 side;
[0085] (5) The intersection of AY1 and the top of the head contour, the intersection of AY2 and the top of the head contour, the minimum ordinate of these three points along the lowest hairline is y2, and a line is drawn along y2 parallel to the x-axis to form the AX2 edge;
[0086] (6) The area formed by the four sides of AX1, AX2, AY1, and AY2 is the frontal A area;
[0087] (7) Extend AX1 and AX2 along the x-axis to the side profile of the head, rotate the extended line downward 15 degrees, and the intersection points with the side profile of the head are By1 and By2. The two adjacent vertices of area A form a quadrilateral with By1 and By2, which is the temporal area B;
[0088] See also Figure 6 and Figure 7 For side-lying posture, the specific method for identifying and dividing the effective temperature measurement area for infants and young children is as follows:
[0089] (1) The opencv library function can be used to identify the eye corners, nose and ear areas, and feature recognition can be used to determine the edge lines of the eye corners, eyebrows and auricles;
[0090] (2) Draw a tangent line BX1 from the point where the auricle connects to the upper side of the head to the eyebrow on the same side, draw a perpendicular line BY2 at half the distance from the tangent point, and draw a perpendicular line BY1 from the canthus of the same side perpendicular to the midline of the eye;
[0091] (3) Draw line BX2 parallel to BX1 at the intersection of BY1 and the hairline on the top of the head; if the hairline cannot be identified, draw line BX2 parallel to BX1 at the distance between the upper edge of the eyebrow and the midpoint of the top of the head;
[0092] (4) The area formed by the intersection of BX1, BX2, BY1, and BY2 is the temporal B area;
[0093] (5) On the side of the temporal area B, extend BX1 and BX2 toward the forehead, and the trapezoidal area formed by the intersection with the edge of the forehead contour is the forehead area A;
[0094] (6) The depression below the earlobe is the C area behind the ear, the height is half the size of the auricle, and the width is set to half the height;
[0095] See also Figure 8 and Fig. 9 ,For prone position, the specific method for identifying and dividing the effective temperature measurement area for infants and young children is as follows:
[0096] (1) The ear area can be identified using the OpenCV library function, and the corners of the eyes, eyebrow edges, auricles, and earlobes can be determined using feature recognition;
[0097] (2) The depression below the earlobe is the C zone behind the ear. Its height is half the size of the auricle and its width is half the height. The outer edge of the C zone behind the ear is an arc line.
[0098] (3) Draw a line BX1 from the point where the auricle connects to the upper side of the head to the edge of the eyebrow on the same side, and draw a perpendicular line BY2 at half the distance from the edge of the eyebrow;
[0099] (4) The area enclosed by the extension of BX1 and BY2 together with the outer edge of the head and the hairline is the temporal area B.
[0100] Step S3: Perform multi-point temperature measurement calculation for the multiple located temperature measurement areas: The three-area division method described in step S2 can be set as follows: for the supine posture, only the forehead area A can be fully identified, and the temperature can be measured by the forehead area A and the temporal area B; for the side-lying posture, only the temporal area B can be fully identified, and the temperature can be measured by the forehead area A, the temporal area B and the area behind the ear C; for the prone posture, only the area behind the ear C can be fully identified, and the temperature can be measured by the temporal area B and the area behind the ear C, see Fig.10 ;
[0101] The specific steps for multi-point temperature measurement calculation in multiple temperature measurement areas are as follows:
[0102] Step S3.1: The obtained forehead area A, temporal area B and behind the ear area C are respectively measured at five points, where the five points are the center point of the area and the temperature measurement points at a position halfway from the upper, lower, left and right edges. The temperatures TA, TB and TC obtained by the temperature measurement are the average temperatures of the five points of the forehead area A, temporal area B and behind the ear area C respectively;
[0103] Step S3.2: Calculate the temperature T0 using different temperature calculation methods according to different sleeping positions:
[0104] (1) If the sleeping position is judged as supine, the temperature is determined based on TA and TB. If |TA-TB| < 0.3°C, T0 = (TA+TB) / 2; otherwise, T0 is equal to the higher temperature of TA and TB.
[0105] (2) If the sleeping position is judged as side-lying, the temperature is measured based on TA, TB, and TC. If |TA-TB| < 0.3°C, T0 = TB; otherwise, T0 is equal to the higher temperature among TA, TB, and TC.
[0106] (3) If the sleeping position is judged as prone, the temperature is measured based on TB and TC. If |TB-TC|<0.3°C, then (TB+TC) / 2; otherwise, T0 is equal to the higher temperature of TB and TC.
[0107] Step S4: Perform environmental compensation correction on the temperature measurement result: measure the ambient temperature T1, use the compensation algorithm based on BP neural network to perform infrared temperature measurement compensation, and obtain the final temperature measurement result T. To ensure the effect of ambient temperature compensation correction, the ambient temperature is best controlled within the range of 15-31°C.
[0108] Step S5: Process the abnormal temperature. The specific processing method is as follows:
[0109] If the final temperature measurement result T is not within the set temperature range of 34-42℃, discard the data and re-measure. If the re-measurement result is still not within the set temperature range, the feedback temperature is wrong.
[0110] If the final temperature measurement result T is within the set temperature range of 34-42°C, but exceeds the alarm threshold, the alarm threshold can be set to 37°C, for example, an alarm is given to the guardian, and the temperature measurement time interval is reduced to half of the original.
[0111] In addition, it should be noted that the shapes and names of the parts and components of the specific embodiments described in this specification may be different, and the above content described in this specification is only an example of the structure of the present invention. All equivalent changes or simple changes made based on the structure, features and principles of the patent concept of the present invention are included in the protection scope of the patent of the present invention. Technicians in the technical field of the present invention can make various modifications or supplements to the specific embodiments described or replace them in a similar manner, as long as they do not deviate from the structure of the present invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.
Claims
1. An intelligent temperature detection method for infants based on multi-region positioning of the head, which is implemented by an intelligent temperature detection system for infants, wherein the intelligent temperature detection system for infants comprises a receiving end, an intelligent analysis module, an infrared temperature measurement module, a camera module, a pan-tilt and a support frame, wherein the pan-tilt is mounted on the support frame, the intelligent analysis module, the infrared temperature measurement module and the camera module are integrated and mounted on the pan-tilt, the camera module is used to shoot infant videos, the intelligent analysis module is used to analyze infant videos and control the infrared temperature measurement module to measure temperature according to the analysis results, and the receiving end is used to receive infant videos and temperature measurement results; Features: The steps of the infant intelligent body temperature detection method based on multi-region positioning of the head are as follows: Initialization: Install the infant intelligent temperature detection system above the infant bed through the support frame. The installation position should ensure that the effective temperature measurement position of the infant can be identified and detected; set the infant's usual sleeping area according to the size and position of the baby bed, that is, the position where parents usually put the infant, and enter the coordinate range into the infant intelligent temperature detection system; Step S1: Detect the sleeping area of the infant and confirm the position of the infant; The specific methods for detecting and confirming the position of infants and young children are: Step S1.1: The camera module takes pictures of the area where infants and young children usually sleep, and the intelligent analysis module uses the Yolov3 model to perform human body recognition to detect whether there are infants and young children in the area; Step S1.2: The infrared temperature measurement module detects the temperature of the identified human body to confirm whether the highest temperature is within the set temperature range, which is 34-42°C; If both of the above conditions are met, it is determined that the infant is in the area, and the process goes to step S2; If it is determined that the infant is not in the area, the camera module's scanning range will be expanded to 3 times the length and width. Repeat step S1, and if it is determined again that the infant is not in the extended area, an alarm is issued; Step S2: Identify the effective temperature measurement area for infants and young children and perform multi-area positioning: The specific steps are: Step S2.1: The head posture of the infant is estimated by the Yolov3 model to determine the sleeping position of the infant. There are three sleeping positions: supine, side-lying, and prone. Step S2.2: Based on the characteristic points of the infant's head, the forehead area A, the temporal area B and the area behind the ear C are identified and delineated. The forehead area A, the temporal area B and the area behind the ear C are the effective areas for measuring the infant's temperature; Step S3: Perform multi-point temperature measurement calculation for the multiple located temperature measurement areas: Settings: For supine posture, only forehead A area can be fully identified, and the temperature can be measured through forehead A area and temporal B area; for side-lying posture, only temporal B area can be fully identified, and the temperature can be measured through forehead A area, temporal B area and behind the ear C area; for prone posture, only behind the ear C area can be fully identified, and the temperature can be measured through temporal B area and behind the ear C area; The specific steps for multi-point temperature measurement calculation in multiple temperature measurement areas are as follows: Step S3.1: The obtained forehead area A, temporal area B and behind the ear area C are respectively measured at five points, where the five points are the center point of the area and the temperature measurement points at a position halfway from the upper, lower, left and right edges. The temperatures TA, TB and TC obtained by the temperature measurement are the average values of the five points of the forehead area A, the temporal area B and behind the ear area C; Step S3.2: Calculate the temperature T0 using different temperature calculation methods according to different sleeping positions: (1) If the sleeping position is judged as supine, the temperature is determined based on TA and TB. If |TA-TB|<0.3℃, T0=(TA+TB) / 2; otherwise, T0 is equal to the higher temperature of TA and TB. (2) If the sleeping position is judged as side-lying, the temperature is measured based on TA, TB and TC. If |TA-TB|<0.3℃, T0=TB; otherwise, T0 is equal to the higher temperature among TA, TB and TC. (3) If the sleeping position is judged as prone, the temperature is measured based on TB and TC. If |TB-TC| < 0.3°C, then (TB+TC) / 2; otherwise, T0 is equal to the higher temperature of TB or TC. Step S4: Perform environmental compensation correction on the temperature measurement result: measure the ambient temperature T1, use the compensation algorithm based on BP neural network to perform infrared temperature measurement compensation, and obtain the final temperature measurement result T; Step S5: Processing abnormal temperature; The method for dealing with abnormal temperature is: If the final temperature measurement result T is not within the set temperature range of 34-42℃, discard the data and re-measure. If the re-measurement result is still not within the set temperature range, the feedback temperature is wrong. If the final temperature measurement result T is within the set temperature range of 34-42°C, but exceeds the alarm threshold, the alarm threshold is set to 37°C, an alarm is given to the guardian, and the temperature measurement time interval is reduced to half of the original.
2. According to claim 1, the method for intelligent temperature detection of infants and young children based on multi-region positioning of the head is characterized by: In step S2.2, the feature points of the infant's head based on the setting include the nose tip, the corners of the eyes, the upper edge of the eyebrows, the earlobes, the hairline and the edge of the forehead, wherein except for the edge of the forehead and the hairline, the other feature points are stable feature points; For infants and young children in the supine position, the specific method for identifying and dividing the effective temperature measurement area is as follows: (1) Use the OpenCV library function to identify the two corners of the eyes and the nose, and calculate the central axis of the head parallel to the nose; (2) Establish a facial region coordinate system and project the face into the xy two-dimensional plane coordinate system. The x-axis corresponds to the line connecting the two eye corners, the y-axis corresponds to the central axis of the head, and the intersection of the two axes is the origin. (3) The two corners of the eyes are divided into horizontal coordinates x1 and x2, and lines are drawn along the y direction parallel to the central axis of the head to form the AY1 and AY2 edges; (4) The line connecting the upper edges of the eyebrows on both sides forms the AX1 edge; (5) The intersection of AY1 and the top of the head contour, the intersection of AY2 and the top of the head contour, the minimum ordinate of these three points along the lowest hairline is y2, and a line is drawn along y2 parallel to the x-axis to form the AX2 edge; (6) The area formed by the four sides of AX1, AX2, AY1, and AY2 is the forehead A area; (7) Extend AX1 and AX2 along the x-axis toward the side profile of the head, rotate the extended line downward 15 degrees, and the intersection points with the side profile of the head are By1 and By2. The two adjacent vertices of area A form a quadrilateral with By1 and By2, which is the temporal area B. For side-lying posture, the specific method for identifying and dividing the effective temperature measurement area for infants and young children is as follows: (1) The opencv library function can be used to identify the corners of the eyes, nose and ear areas, and feature recognition can be used to determine the edge lines of the corners of the eyes, eyebrows and auricles; (2) Draw a tangent line BX1 from the point where the auricle connects to the upper side of the head to the eyebrow on the same side, draw a perpendicular line BY2 at half the distance from the tangent point, and draw a perpendicular line BY1 from the canthus of the same side perpendicular to the midline of the eye; (3) Draw line BX2 parallel to BX1 at the intersection of BY1 and the hairline on the top of the head. If the hairline cannot be identified, draw line BX2 parallel to BX1 at the distance between the upper edge of the eyebrow and the midpoint of the top of the head. (4) The area formed by the intersection of BX1, BX2, BY1, and BY2 is the temporal B area; (5) On the side of the temporal area B, extend BX1 and BX2 toward the forehead, and the trapezoidal area formed by the intersection with the edge of the forehead contour is the forehead area A; (6) The depression below the earlobe is the postauricular C area, which has a height of half the size of the auricle and a width of half the height; For prone posture, the specific method for identifying and dividing the effective temperature measurement area for infants and young children is as follows: (1) The ear area can be identified using the OpenCV library function, and the corners of the eyes, eyebrow edges, auricles, and earlobes can be determined using feature recognition; (2) The depression below the earlobe is the C area behind the ear. Its height is half the size of the auricle, and its width is set to half the height. The outer edge of the C area behind the ear is an arc line. (3) Draw a line BX1 from the point where the auricle connects to the upper side of the head to the edge of the eyebrow on the same side, and draw a perpendicular line BY2 at half the distance from the edge of the eyebrow; (4) The area enclosed by the extension of BX1 and BY2 together with the outer edge of the head and the hairline is the temporal zone B.
Citation Information
Patent Citations
Sleep monitoring method, device and system and electronic equipment
CN108209871A
Process for the contactless determination of the skin temperature and device for carrying out the process
CN109813461A
Anti-drop early warning method and system for sleeping infants based on motion detection
CN109984750A
Multi-purpose video monitoring camera
US20210290105A1