Attitude detection method and device and readable storage medium

Abstract

The invention provides a posture detection method, which is used for detecting the posture of a human body and comprises the following steps: acquiring a first infrared thermal imaging image of a target scene; extracting first pixel points corresponding to a human body in the first infrared thermal imaging image to form a first human body image area; acquiring a second infrared thermal imaging image of the target scene; extracting second pixel points corresponding to a human body in the second infrared thermal imaging image to form a second human body image area; when the overlapping degree of the first human body image area and the second human body image area is lower than a preset overlapping degree, the length-width ratio of a minimum circumscribed rectangle corresponding to a minimum circumscribed ellipse of the second human body image area and the included angle between the long axis of the minimum circumscribed ellipse and the preset vertical direction are obtained; and judging the posture of the human body according to the length-width ratio and the included angle. The invention further provides a posture detection device and a readable storage medium.

Description

Technical Field

[0001] The present application relates to a posture detection method, device and readable storage medium. Background Art

[0002] In a home environment, users may fall for various reasons, which poses a risk to the health of users (especially elderly users). Therefore, a fall detection system can be used for home safety protection.

[0003] However, due to personal privacy protection and security issues, the high resolution of traditional cameras makes them unsuitable for use in highly private home scenarios (such as bathrooms, bedrooms, etc.). Summary of the invention

[0004] The present application provides a posture detection method for detecting a human body posture, comprising the following steps: Acquire a first infrared thermal imaging image of a target scene, wherein the first infrared thermal imaging image includes a plurality of first pixel points; Extracting a first pixel point corresponding to a human body in the first infrared thermal imaging image to form a first human body image area; Acquire a second infrared thermal imaging image of the target scene, wherein the second infrared thermal imaging image includes a plurality of second pixel points; Extracting second pixel points corresponding to a human body in the second infrared thermal imaging image to form a second human body image area; and When the overlap between the first human body image area and the second human body image area is lower than a preset overlap, the aspect ratio of the minimum circumscribed rectangle corresponding to the minimum circumscribed ellipse of the second human body image area and the angle between the major axis of the minimum circumscribed ellipse and a preset vertical direction are obtained, and the human body posture is determined based on the aspect ratio and the angle.

[0005] The above posture detection method effectively protects personal privacy because it obtains infrared thermal imaging images. The posture of the human body is determined based on the characteristics of the minimum circumscribed ellipse of the second human body image area, and the aspect ratio of the minimum circumscribed rectangle corresponding to the minimum circumscribed ellipse and the angle between the major axis of the minimum circumscribed ellipse and the preset vertical direction are used as the basis for determination, thereby improving the accuracy of the posture detection method and helping to reduce the probability of false alarms.

[0006] A second aspect of the present application provides a posture detection device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the posture detection method described above when executing the computer program.

[0007] For the above-mentioned posture detection device, by executing the above-mentioned posture detection method, all the beneficial effects of the above-mentioned posture detection method can be achieved.

[0008] A third aspect of the present application provides a readable storage medium storing a computer program, wherein the computer program implements the posture detection method as described above when executed.

[0009] The above-mentioned readable storage medium is integrated with the above-mentioned posture detection method, and all the beneficial effects of the above-mentioned posture detection method can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 Schematic diagram of a module of a posture detection device according to an embodiment of the present application.

[0011] Figure 2 This is a flow chart of a posture detection method according to an embodiment of the present application.

[0012] Figure 3 It is a schematic diagram of a first infrared thermal imaging image obtained in step S1 of the posture detection method of an embodiment of the present application and a processing process of the first infrared thermal imaging image.

[0013] Figure 4 Schematic diagram of the processing process of the first infrared thermal imaging image in step S2 of the posture detection method in an embodiment of the present application.

[0014] Figure 5 Schematic diagram of the minimum circumscribed ellipse and the minimum circumscribed rectangle of the second human image area.

[0015] Main component symbols Posture detection device: 100 Thermal imaging cameras: 10 Filters: 20 - Processor: 30 Memory: 40 Infrared thermal imaging images: 1, 2, 3 First initial human image area: 31 First human image area: 32 Minimum circumscribed ellipse: 33 The following specific implementation methods will further illustrate the present application in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0016] The embodiments of the present application provide a posture detection method and device for detecting a user's posture to determine whether the user has fallen, and are suitable for home scenes with high privacy (such as bathrooms, bedrooms, etc.).

[0017] See also Figure 1The posture detection device 100 of the embodiment of the present application includes a thermal imaging camera 10, a filter 20, a processor 30 and a memory 40. The thermal imaging camera 10 is used to obtain infrared thermal imaging images, and the filter 20 is used to remove noise in the infrared thermal imaging images, and retain the edge contours of the images to improve the quality of the images. The memory 40 stores a computer program, and when the computer program is read by the processor 30, the posture detection device 100 executes a posture detection method to detect the posture of the human body based on the infrared thermal imaging image taken by the thermal imaging camera 10, and determine whether the human body has fallen. In other embodiments, the posture detection device 100 also includes an alarm, which is used to sound an alarm when a human body falls, so as to promptly remind the user. Please refer to Figure 2 The posture detection method provided in the embodiment of the present application is applied to the above-mentioned posture detection device 100, and includes the following steps: S1, acquiring a first infrared thermal imaging image of a target scene, where the first infrared thermal imaging image includes a plurality of first pixel points.

[0018] S2, extracting a first pixel point corresponding to a human body in the first infrared thermal imaging image to form a first human body image area.

[0019] S3, acquiring a second infrared thermal imaging image of the target scene, where the second infrared thermal imaging image includes a plurality of second pixel points.

[0020] S4, extracting second pixel points corresponding to the human body in the second infrared thermal imaging image to form a second human body image area.

[0021] S5, when the overlap between the first human body image area and the second human body image area is lower than a preset overlap, obtain the aspect ratio of the minimum circumscribed rectangle corresponding to the minimum circumscribed ellipse of the second human body image area, and the angle between the major axis of the minimum circumscribed ellipse and the preset vertical direction, and determine the human body posture according to the aspect ratio and the angle.

[0022] In step S1, a first infrared thermal imaging image can be acquired through the thermal imaging camera 10. Infrared thermal imaging mainly uses photoelectric technology to detect infrared specific band signals of thermal radiation of objects, converts the signals into images and graphics that can be distinguished by human vision, and can further calculate the temperature value of each pixel in the image. Therefore, in the first infrared thermal imaging image, each first pixel corresponds to a temperature value. In this embodiment, the thermal imaging camera 10 has a low resolution, so that the clarity of the acquired image is not high, which is beneficial to protecting the privacy of the user. In addition, the use of a thermal imaging camera 10 with a low resolution is conducive to reducing costs.

[0023] Furthermore, in order to minimize the influence of the external environment, a Gaussian filter or a median filter may be used to perform a first denoising process on the first infrared thermal imaging image, thereby filtering out noise in the infrared thermal imaging data and effectively improving the image quality. In this embodiment, a first denoising process is performed on the first infrared thermal imaging image by a filter 20, wherein the filter 20 is a Gaussian filter or a median filter.

[0024] In some embodiments, step S1 further includes performing a second denoising process on the first infrared thermal imaging image after acquiring the first infrared thermal imaging image, specifically including the following steps: Step S11, extracting a first pixel point whose temperature value is within a preset temperature range in a first infrared thermal imaging image.

[0025] Step S12, normalizing the temperature value corresponding to the first pixel point whose temperature value is within the preset temperature range, enhancing the contrast between the bright part and the dark part in the first infrared thermal imaging image, thereby improving the clarity of the first infrared thermal imaging image.

[0026] Step S13, retaining the edge information in the first infrared thermal imaging image that has undergone the normalization process and performing smoothing on the image portion other than the edge information.

[0027] In step S11, the preset temperature range is set according to the human body temperature. The normal human body temperature range is approximately 36°C to 37.7°C, and the preset temperature range will slightly exceed the normal human body temperature range, such as the preset temperature range may be 20°C to 40°C. Since the human body temperature may change with age, gender, activity, weather, medication and other factors, the preset temperature range may change according to the change of the human body temperature.

[0028] In step S13, the edge information in the first infrared thermal imaging image refers to the part of the first infrared thermal imaging image that reflects the most significant temperature change on the surface of the object, and also refers to the boundary between the area with a gentle grayscale change and the area with a rapid grayscale change in the image. Infrared thermal imaging images are imaged based on the thermal radiation emitted by the object, and the temperature change on the surface of the object will cause differences in infrared radiation, and these differences are manifested as edge information in the infrared thermal imaging image. For example, when there are living things in the scene, due to the different surface temperatures of the living things and the surrounding non-living matter, there is a difference in infrared radiation emitted by the living things and the surrounding non-living matter, so that there is a grayscale change between the image part corresponding to the living thing and the image part corresponding to the surrounding non-living thing in the infrared thermal imaging image, and then there is a boundary between the image of the living thing and the surrounding non-living thing, and the pixels corresponding to the boundary form the edge information.

[0029] Specifically, retaining edge information in the first infrared thermal imaging image can be achieved by using a filter 20, and the filter 20 is an edge preserving filter. When encountering the edge of the first infrared thermal imaging image, the edge preserving filter tries to eliminate its smoothness, and at the same time makes the general area of ​​the first infrared thermal imaging image show smoothness, that is, retaining the edge information in the first infrared thermal imaging image, and smoothing the image part other than the edge information in the first infrared thermal imaging image, thereby improving the clarity and visibility of the first infrared thermal imaging image. The edge preserving filter includes a bilateral filter, a guided filter, and a weighted least squares filter, and their working principles are different. According to different application scenarios, different edge preserving filters can be selected to process the first infrared thermal imaging image. For example, the bilateral filter is suitable for image processing tasks that need to retain edge details, while the guided filter and the weighted least squares filter are suitable for more complex image processing tasks. The selection of a specific filter depends on the specific application requirements and image characteristics.

[0030] See also Figure 3 , Figure 3 It includes three infrared thermal imaging images 1, 2, and 3. Infrared thermal imaging image 1 corresponds to the schematic diagram of the first initial image that has not been processed in the above step S1, infrared thermal imaging image 2 corresponds to the schematic diagram of the first normalized image obtained after normalizing the temperature value in the above step S1 (i.e., processed by step S12), and infrared thermal imaging image 3 corresponds to the schematic diagram of the first infrared thermal imaging image obtained after retaining the edge information in the above step S (i.e., processed by step S13). It can be seen from the figure that the number of gaps in the edge contour line of the human body in the infrared thermal imaging image 2 is less than the number of gaps in the edge contour line of the human body in the infrared thermal imaging image 1, and the number of gaps in the edge contour line of the human body in the infrared thermal imaging image 3 is less than the number of gaps in the edge contour line of the human body in the infrared thermal imaging image 2, which indicates that the clarity of the infrared thermal imaging image 3 is greater than that of the infrared thermal imaging image 2, and the clarity of the infrared thermal imaging image 2 is greater than that of the infrared thermal imaging image 1. It can be seen that after the first infrared thermal imaging image is normalized, the edge information is retained, and the smoothing process is performed, the clarity of the first infrared thermal imaging image is improved, which is convenient for effectively identifying the human body in the first infrared thermal imaging image.

[0031] That is, in this embodiment, the thermal imaging camera 10 continuously acquires infrared thermal imaging images, and the infrared thermal imaging images acquired by the thermal imaging camera 10 are sequentially used as the first infrared thermal imaging image, the second infrared thermal imaging image, the third infrared thermal imaging image, etc. in chronological order. Correspondingly, the filter 20 also continuously performs denoising processing on each acquired infrared thermal imaging image and retains the edge information of the image.

[0032] It should be noted that in the above step S1, the infrared thermal imaging image initially acquired may be directly subjected to the second denoising process (ie, the first denoising process is omitted), or may be subjected to the first denoising process and then the second denoising process, so as to improve the image quality.

[0033] In step S2, it mainly includes: S21, extracting a portion of first pixel points corresponding to a human body in the first infrared thermal imaging image to form a first initial human body image region; S22, comparing the temperature value distribution corresponding to each first pixel point in the first initial human body image area with the foreground area temperature model, when the temperature value distribution corresponding to each first pixel point in the first initial human body image area is consistent with the foreground area temperature model, reducing the first initial human body image area based on a connected components algorithm to obtain the first human body image area.

[0034] In step S21, the step of extracting part of the first pixel points corresponding to the human body in the first infrared thermal imaging image to form a first initial human body image area includes: extracting part of the first pixel points corresponding to the human body in the first infrared thermal imaging image to form a first initial human body image area based on computer vision (Computer Vision, CV) and machine learning (Machine Learning, ML), or based on deep learning (Deep Learning, DL).

[0035] Among them, the method of extracting the first initial human image area in the first infrared thermal imaging image based on CV and ML can be implemented by a support vector machine (SVM) classifier according to the histogram gradient (HoG) feature, and the method of extracting the first initial human image area in the first infrared thermal imaging image based on DL can be implemented by a YOLO (You Only Look Once) detector. The method based on CV and ML requires less calculation and has low hardware requirements, and is suitable for simple environments. The method based on DL requires a large amount of calculation and has high hardware requirements, but the execution process is simpler when using this method, and is suitable for complex environments.

[0036] See also Figure 4 , Figure 4In the process of extracting the first pixel points corresponding to the human body in the first infrared thermal imaging image 3 to form the first human body image area 32, it can be seen that the area of ​​the first human body image area is smaller than the area of ​​the first initial human body image area. Since the method based on CV and ML and the method based on DL are limited by the algorithm when extracting the human body area in the first infrared thermal imaging image 3, the extracted first initial human body image area 31 is relatively large. Therefore, this embodiment adopts the connected body algorithm to reduce the first initial human body image area 31 to obtain the first human body image area 32. The algorithm can more accurately extract the first human body image area 32 in the first infrared thermal imaging image 3, thereby improving the accuracy of the posture detection method.

[0037] In step S22, the foreground area temperature model needs to be obtained through the following steps: Acquire multiple infrared thermal imaging images of the environment in which the human body is located; According to temperature values ​​corresponding to a plurality of pixel points in a plurality of infrared thermal imaging images, a temperature distribution model corresponding to each pixel point is established by using a Gaussian Mixture Model (GMM), wherein the temperature distribution model includes a foreground area temperature model and a background area temperature model.

[0038] Since the temperature model of the foreground area has the same temperature distribution as the human body, and the temperature model of the background area has the same temperature distribution as the environment in which the human body is located, by comparing the temperature value distribution corresponding to each first pixel point in the first initial human image area with the temperature model of the foreground area, it is ensured that the first initial human image area contains a human body, thereby improving the accuracy of the posture detection method.

[0039] After the first human body image area is extracted through step S21 and step S22, it is necessary to fix the preset temperature and stop acquiring the foreground area temperature model to ensure that in the subsequent steps, the second human body image area in the second infrared thermal imaging image is extracted under the same environment and conditions to reduce system errors.

[0040] The above step S2 is executed by the processor 30 .

[0041] In some embodiments, step S3 further includes processing the second infrared thermal imaging image after acquiring the second infrared thermal imaging image, and the processing process of the second infrared thermal imaging image is the same as the processing process of the first infrared thermal imaging image in the above step S1.

[0042] In some embodiments, when step S4 extracts part of the second pixel points corresponding to the human body in the second infrared thermal imaging image to form a second human body image area, the same processing process as that of extracting part of the first pixel points corresponding to the human body in the first infrared thermal imaging image to form a first human body image area in the above step S2 can be adopted.

[0043] In step S5, it mainly includes: S51, by comparing the first human body image region with the second human body image region, when the overlap between the first human body image region and the second human body image region is lower than a preset overlap, directly outputting the second human body image region; When the degree of overlap between the first human body image region and the second human body image region is higher than the preset degree of overlap, a first part of the first human body image region that is different from the second human body image region is removed, a second part of the second human body image region that is different from the first human body image region is extracted, and the second part is merged with the first human body image region from which the first part is removed, so as to update the first human body image region and output it as the second human body image region.

[0044] S52, detecting whether there is an area in the second infrared thermal imaging image that has a similar color scale distribution to that of the first human body image area, and when it is detected that there is an area in the second infrared thermal imaging image that has a similar color scale distribution to that of the first human body image area, combining the similar area with the second human body image area and outputting the second human body image area.

[0045] S53, obtaining the aspect ratio of the minimum circumscribed rectangle corresponding to the minimum circumscribed ellipse of the second human body image region, and the angle between the major axis of the minimum circumscribed ellipse and a preset vertical direction.

[0046] S54, determining the posture of the human body according to the aspect ratio and the angle, including the following steps: Determine whether the included angle is within a preset angle range and whether the aspect ratio is less than a preset ratio. When the included angle is outside the preset angle range, determine that the human body is in a normal state; When the included angle is within the preset angle range and the aspect ratio is less than a preset ratio, it is determined that the human body is in a state of possible falling; When the time that the human body is in the possible falling state exceeds a preset time length, it is determined that the human body has fallen and an alarm is issued.

[0047] In step S51, the overlap is used to measure the number of pixels with the same grayscale in the first human image region and the second human image region. When the number of pixels with the same grayscale in the first human image region and the second human image region is large, the overlap between the first human image region and the second human image region is high; when the number of pixels with the same grayscale in the first human image region and the second human image region is small, the overlap between the first human image region and the second human image region is low.

[0048] Since the grayscale of each pixel in the human body image area is formed according to the temperature value corresponding to each pixel, when the grayscale of the first pixel in the first human body image area is the same as the grayscale of the second pixel in the second human body image, the temperature value corresponding to the first pixel in the first human body image area is the same as the temperature value corresponding to the second pixel in the second human body image; when the grayscale of the first pixel in the first human body image area is the same as the grayscale portion of the second pixel in the second human body image, the temperature value corresponding to the first pixel in the first human body image area is also different from the temperature value corresponding to the second pixel in the second human body image.

[0049] The preset overlap is set according to the number of pixels with the same grayscale in the first human image area and the second human image area. When the overlap between the first human image area and the second human image area is lower than the preset overlap, it indicates that the number of pixels with the same grayscale in the first human image area and the second human image area is small, the temperature values ​​corresponding to most of the first pixels in the second human image are different from the temperature values ​​corresponding to the second pixels in the second human image, and there is a temperature distribution in the second human image area that is different from that in the first human image area. At this time, the second human image area is directly output.

[0050] When the degree of overlap between the first human body image area and the second human body image area is higher than a preset degree of overlap, it indicates that there are many pixels with the same grayscale in the first human body image area and the second human body image area, the temperature values ​​corresponding to most of the first pixels in the second human body image are the same as the temperature values ​​corresponding to the second pixels in the second human body image, and the human body temperature distribution in the second human body image area is basically the same as the temperature distribution in the first human body image area. At this time, the first human body image area is updated and output as the second human body image area.

[0051] The above step S51 compares the first human body image area with the second human body image area, directly outputs the second human body image area when the overlap is low, and updates the first human body image area and outputs it as the second human body image area when the overlap is high, thereby ensuring that the human body posture in the second human body image area will not be repeatedly detected when the second human body image area is subsequently detected.

[0052] In step S52, by detecting whether there is an area in the second infrared thermal imaging image that has a similar color scale distribution to that of the first human body image area, that is, detecting whether there is an area in the second infrared thermal imaging image in which the temperature distribution composed of temperature values ​​corresponding to multiple pixels is the same as the temperature distribution composed of temperature values ​​corresponding to multiple pixels in the first human body image area, it is ensured that the output second human body image area contains all the human body areas in the second infrared thermal imaging image, thereby improving the accuracy of the posture detection method.

[0053] If no region similar to the color scale distribution of the first human image region is detected in the second infrared thermal imaging image in step S52, it means that no human image region similar to the first human image region is present in the second infrared thermal imaging image. When the duration of not detecting the presence of a region similar to the color scale distribution of the first human image region in the second infrared thermal imaging image exceeds the preset life cycle, it indicates that the human body in the first human image region has left the current scene, and at this time, the determination of the human body posture in the first human image region is stopped. The preset life cycle is related to the size of the scene in which the human body is located. When the scene is large, the preset life cycle is long, and when the scene is small, the preset life cycle is short.

[0054] See also Figure 5 , Figure 5 The ellipse in the figure represents the minimum circumscribed ellipse of a human image area. The method for obtaining the minimum circumscribed ellipse is the same as Figure 4 The acquisition method of the minimum circumscribed ellipse 33 of the first human image region 32 is the same, and it needs to be acquired close to the edge contour line of the first human image region 32. Figure 5 The rectangle in Figure 5 In this embodiment, the preset angle range in step S54 is 45° to 145°, the preset ratio is 1.25, and the preset duration is 1 second. Figure 5 It is a schematic diagram of the minimum circumscribed ellipse of the second human body image area in a plane rectangular coordinate system when the human body is in a state where the human body may fall, and the y-axis of the plane rectangular coordinate system is a preset vertical direction. Figure 5 The angle θ in is greater than 45° and less than 145°, and is within the preset angle range. The length of the minimum circumscribed rectangle is the short axis of the minimum circumscribed ellipse, and the width of the minimum circumscribed rectangle is the long axis of the minimum circumscribed ellipse, and the aspect ratio of the minimum circumscribed rectangle is less than 1.25. When the angle θ is outside the preset angle range, the length of the minimum circumscribed rectangle is the long axis of the minimum circumscribed ellipse, and the width of the minimum circumscribed rectangle is the short axis of the minimum circumscribed ellipse. If the human body is in a state of possible fall for more than 1 second, it is determined that the human body has fallen and an alarm is issued.

[0055] The angle is the angle between the major axis of the minimum circumscribed ellipse and the preset vertical direction, and the preset vertical direction is the direction of the plumb line in the scene where the human body is located. When the position of the posture detection device in the scene changes, the image of the human body in the infrared thermal imaging image acquired by the device changes accordingly, resulting in changes in the preset angle range, preset ratio and preset duration. Therefore, this application does not limit the preset angle range, preset ratio and preset duration, and sets different preset angle ranges, preset ratios and preset durations according to the changes in the position of the posture detection device in the scene.

[0056] The above step S5 is executed by the processor 30 .

[0057] In summary, the resolution of the infrared thermal imaging image obtained in the posture detection method of the present application is relatively low, and the human body contour can be displayed for posture determination, which is beneficial to protecting personal safety. On this basis, since the details of the human body are not displayed, it is beneficial to ensure the privacy security of home scenes with high privacy. The posture detection method of the present application also determines whether the human body falls based on the characteristics of the minimum circumscribed ellipse of the second human body image area, and adopts a three-level alarm mechanism, wherein the first state refers to the human body being in a normal state, and the second state refers to the human body being in a state where it may fall. When the time when the human body is in a state where it may fall exceeds the preset time, the second state is changed to the third state, and the third state refers to the human body falling and an alarm is issued. Through the above steps, the accuracy and stability of the detection are enhanced, which is conducive to reducing the probability of false alarms.

[0058] An embodiment of the present application further provides a non-volatile computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of any of the above-mentioned posture detection methods when executed by a processor.

[0059] When the above-mentioned posture detection method is implemented in the form of a separate product and sold or used, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned implementation method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0060] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the sample processing system / sample processing device, and uses various interfaces and lines to connect various parts of the entire sample processing system / sample processing device.

[0061] The memory is used to store the computer program and / or module, and the processor realizes various functions of the sample processing system / sample processing device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.). In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0062] Those skilled in the art should recognize that the above embodiments are only used to illustrate the present application and are not intended to be limiting of the present application. As long as they are within the spirit and scope of the present application, appropriate changes and modifications to the above embodiments are within the scope of protection claimed in the present application.

Claims

1. A posture detection method, characterized in that: Used to detect human posture, including the following steps: Acquire a first infrared thermal imaging image of a target scene, wherein the first infrared thermal imaging image includes a plurality of first pixel points; Extracting a first pixel point corresponding to a human body in the first infrared thermal imaging image to form a first human body image area; Acquire a second infrared thermal imaging image of the target scene, wherein the second infrared thermal imaging image includes a plurality of second pixel points; Extracting second pixel points corresponding to a human body in the second infrared thermal imaging image to form a second human body image area; as well as When the overlap between the first human body image area and the second human body image area is lower than a preset overlap, the aspect ratio of the minimum circumscribed rectangle corresponding to the minimum circumscribed ellipse of the second human body image area and the angle between the major axis of the minimum circumscribed ellipse and a preset vertical direction are obtained, and the human body posture is determined based on the aspect ratio and the angle.

2. The posture detection method according to claim 1, characterized in that: The step of extracting some first pixel points corresponding to the human body in the first infrared thermal imaging image to form a first human body image area includes: extracting some first pixel points corresponding to the human body in the first infrared thermal imaging image to form a first initial human body image area, and when the temperature value distribution corresponding to each first pixel point in the first initial human body image area is consistent with the foreground area temperature model, reducing the first initial human body image area based on a connected body algorithm to obtain the first human body image area; The step of extracting part of the second pixel points corresponding to the human body in the second infrared thermal imaging image to form a second human body image area includes: extracting part of the second pixel points corresponding to the human body in the second infrared thermal imaging image to form a second initial human body image area, and when the temperature value distribution corresponding to each second pixel point in the second initial human body image area is consistent with the foreground area temperature model, the second initial human body image area is reduced based on a connected body algorithm to obtain the second human body image area.

3. The posture detection method as claimed in claim 2, characterized in that: The step of extracting part of the first pixel points corresponding to the human body in the first infrared thermal imaging image to form a first initial human body image area comprises: extracting part of the first pixel points corresponding to the human body in the first infrared thermal imaging image to form a first initial human body image area based on computer vision and machine learning, or based on deep learning; The step of extracting part of the second pixel points corresponding to the human body in the second infrared thermal imaging image to form a second initial human body image area includes: extracting part of the second pixel points corresponding to the human body in the second infrared thermal imaging image to form a second initial human body image area based on computer vision and machine learning, or based on deep learning.

4. The posture detection method according to claim 1, characterized in that: When the degree of overlap between the first human body image area and the second human body image area is higher than the preset degree of overlap, a first part of the first human body image area that is different from the second human body image area is removed, a second part of the second human body image area that is different from the first human body image area is extracted, and the second part is merged with the first human body image area from which the first part is removed, so as to update the first human body image area and output it as the second human body image area.

5. The posture detection method according to claim 1, characterized in that: Before the step of obtaining the aspect ratio of the minimum circumscribed rectangle corresponding to the minimum circumscribed ellipse of the second human image area, and the angle between the major axis of the minimum circumscribed ellipse and the preset vertical direction, the step further includes: When it is detected that there is a region in the second infrared thermal imaging image that has a color scale distribution similar to that of the first human image region, the similar region is combined with the second human image region and the second human image region is output.

6. The posture detection method according to claim 1, characterized in that: The step of determining the posture of a human body according to the aspect ratio and the angle comprises: When the angle is outside the preset angle range, it is determined that the human body is in a normal state; When the included angle is within the preset angle range and the aspect ratio is less than a preset ratio, it is determined that the human body is in a state of possible falling; When the time that the human body is in the possible falling state exceeds a preset time, it is determined that the human body has fallen and an alarm is issued.

7. The posture detection method according to claim 6, characterized in that: The preset angle range is 45°~145°; or The preset ratio is 1.25; or The preset duration is 1 second.

8. The posture detection method according to claim 1, characterized in that: Each of the first pixel points and each of the second pixel points corresponds to a temperature value, After the step of acquiring the first infrared thermal imaging image, the method further includes: extracting part of the first pixel points whose temperature values ​​are within a preset temperature range in the first infrared thermal imaging image, normalizing the temperature values ​​corresponding to the part of the first pixel points whose temperature values ​​are within the preset temperature range, retaining edge information in the first infrared thermal imaging image, and then extracting the first human body image in the first infrared thermal imaging image; After the step of acquiring the second infrared thermal imaging image, it also includes: extracting some of the second pixel points in the second infrared thermal imaging image whose temperature values ​​are within a preset temperature range, normalizing data composed of temperature values ​​corresponding to some of the second pixel points whose temperature values ​​are within the preset temperature range, retaining edge information in the second infrared thermal imaging image, and then extracting the first human body image in the first infrared thermal imaging image.

9. A posture detection device, characterized in that: The posture detection device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the posture detection method according to any one of claims 1 to 8 is implemented.

10. A readable storage medium storing a computer program, characterized in that: When the computer program is executed, the posture detection method according to any one of claims 1 to 8 is implemented.

Images