An infrared thermal image processing method, device, equipment and medium

Abstract

This application discloses a method, apparatus, device, and medium for processing infrared thermal images, belonging to the field of image processing technology. The method includes: preprocessing a target infrared thermal image to obtain a preprocessed infrared thermal image; the target infrared thermal image contains at least one object to be measured; identifying various parts of the target object to be measured in the preprocessed infrared thermal image to obtain target identification information; the target object to be measured is any object to be measured in the target infrared thermal image; determining the central axis of the target object to be measured in the preprocessed infrared thermal image based on the target identification information to obtain the target central axis, and generating a region of interest (ROI) in the preprocessed infrared thermal image based on the target identification information; and performing image analysis on the target object to be measured in the target infrared thermal image based on the target central axis and the ROI. This method can fill the current technological gap of not utilizing infrared thermal imaging technology for image analysis.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method, apparatus, device, and medium for processing infrared thermal images. Background Technology

[0002] Infrared thermal imaging (also known as infrared thermography or thermal imaging) is a non-invasive detection technology that uses infrared detectors to capture infrared radiation from an object's surface and convert it into a temperature distribution image. In daily life, infrared thermal imaging is mostly used for screening people's body temperature. However, there are currently no patents related to image analysis using infrared thermal imaging, which limits its in-depth application in real life. This issue urgently needs to be addressed by those skilled in the art. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide a method, apparatus, device, and medium for processing infrared thermal images, thereby filling the technological gap in image analysis that currently does not utilize infrared thermal imaging technology. The specific solution is as follows: To address the aforementioned technical problems, this invention provides a method for processing infrared thermal images, comprising: The infrared thermal image of the target is preprocessed to obtain a preprocessed infrared thermal image; the target infrared thermal image contains at least one object to be measured. The target object to be tested is identified by identifying various parts of the target in the preprocessed infrared thermal image to obtain target identification information; the target object to be tested is any human body in the target infrared thermal image; Based on the target recognition information, the central axis of the target object to be measured is determined in the preprocessed infrared thermal image to obtain the target central axis, and the target region of interest is generated in the preprocessed infrared thermal image based on the target recognition information. Image analysis is performed on the target object in the infrared thermal image of the target based on the target's central axis and the target's region of interest.

[0004] Preferably, the preprocessing of the target infrared thermal image to obtain a preprocessed infrared thermal image includes: The target infrared thermal image is subjected to non-uniformity correction, temperature normalization, noise suppression, and background attenuation processing to obtain the preprocessed infrared thermal image.

[0005] Preferably, the step of identifying various parts of the target object in the preprocessed infrared thermal image to obtain target identification information includes: The target identification information is obtained by using a posture key point detection model and / or a structural template matching method to identify various parts of the target object in the preprocessed infrared thermal image.

[0006] Preferably, the step of determining the central axis of the target object in the preprocessed infrared thermal image based on the target identification information to obtain the target central axis includes: Based on the key point fitting algorithm and / or principal direction analysis method and / or statistical regression method, and according to the target identification information, the central axis of the target object to be measured is determined in the preprocessed infrared thermal image to obtain the target central axis.

[0007] Preferably, generating a region of interest for a target in the preprocessed infrared thermal image based on the target identification information includes: The target region of interest is generated in the preprocessed infrared thermal image based on the structural proportion rule generation algorithm and / or temperature gradient algorithm and / or region clustering algorithm and / or rule-guided region generation algorithm, according to the target identification information.

[0008] Preferred options also include: If the number of the target regions of interest is greater than one, then the temperature value corresponding to each target region of interest is calculated; The average temperature of all the regions of interest is determined based on the temperature value corresponding to each of the target regions of interest and the total number of all the target regions of interest, thus obtaining the target average value; If there is a region of interest with a temperature value greater than the average value of the target among the multiple regions of interest, then the region of interest with a temperature value greater than the average value of the target is marked to obtain a marked region of interest; If the area of ​​the marked region of interest is larger than a preset area, and the duration for which the temperature value corresponding to the marked region of interest is greater than the target average value exceeds a preset duration, then the target region of interest is determined to be an abnormal region of interest, and a warning message is issued.

[0009] Preferred options also include: The infrared thermal images corresponding to the target object to be tested are collected according to a preset cycle to obtain multiple frames of infrared thermal images; The multi-frame infrared thermal images are processed, and the target region of interest corresponding to the target object under test in the multi-frame infrared thermal images is obtained to obtain the target set; The temperature information corresponding to all regions of interest of the targets in the target set is determined, and the temperature information corresponding to all regions of interest of the targets in the target set is sorted according to the time order of the infrared thermal images to obtain time series data; The time series data is used to evaluate the state changes of the target object under test.

[0010] To address the aforementioned technical problems, the present invention also provides an infrared thermal image processing apparatus, comprising: An image preprocessing module is used to preprocess the target infrared thermal image to obtain a preprocessed infrared thermal image; the target infrared thermal image contains at least one object to be measured; The information recognition module is used to identify various parts of the target object in the preprocessed infrared thermal image to obtain target recognition information; the target object is any one of the target objects in the target infrared thermal image. The region generation module is used to determine the central axis of the target object to be measured in the preprocessed infrared thermal image based on the target identification information, obtain the target central axis, and generate the target region of interest in the preprocessed infrared thermal image based on the target identification information. The image analysis module is used to perform image analysis on the target object in the infrared thermal image of the target based on the target's central axis and the target's region of interest.

[0011] To address the aforementioned technical problems, the present invention also provides an electronic device, comprising: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of a method for processing an infrared thermal image as disclosed above.

[0012] To address the aforementioned technical problems, the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the infrared thermal image processing method disclosed above.

[0013] Beneficial Effects: In the infrared thermal image processing method provided by this invention, the target infrared thermal image is first preprocessed to obtain a preprocessed infrared thermal image; wherein the target infrared thermal image contains at least one object to be measured; then, various parts of the target object to be measured in the preprocessed infrared thermal image are identified to obtain target identification information; wherein the target object to be measured is any object to be measured in the target infrared thermal image; then, the central axis of the target object to be measured is determined in the preprocessed infrared thermal image based on the target identification information to obtain the target central axis, and a target region of interest is generated in the preprocessed infrared thermal image based on the target identification information; finally, image analysis is performed on the target object to be measured in the target infrared thermal image based on the target central axis and the target region of interest. This method can use infrared thermal imaging technology to identify various parts of the object to be measured and determine the central axis and region of interest of the object to be measured based on the identified information. Since the central axis and region of interest of the object to be measured can be used for image analysis operations such as medical image annotation, symmetry analysis, and state assessment of the object to be measured, the method provided by this invention can fill the technical gap of not currently utilizing infrared thermal imaging technology for image analysis.

[0014] Correspondingly, the infrared thermal image processing apparatus, device, and medium provided by the present invention also have the above-mentioned beneficial effects. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0016] Figure 1 A flowchart illustrating an infrared thermal image processing method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram showing the connection between a medical infrared thermal imaging device and a host computer. Figure 3 This is a flowchart of an embodiment of the present invention for image analysis of multiple objects under test in an infrared thermal image; Figure 4 This is a flowchart illustrating a key point fitting algorithm provided in an embodiment of the present invention, which determines the central axis of the target object in a preprocessed infrared thermal image based on target recognition information. Figure 5 A flowchart illustrating another infrared thermal image processing method provided in an embodiment of the present invention; Figure 6A flowchart illustrating yet another infrared thermal image processing method provided in an embodiment of the present invention; Figure 7 This is a structural diagram of an infrared thermal image processing device provided in an embodiment of the present invention; Figure 8 This is a structural diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 , Figure 1 A flowchart of an infrared thermal image processing method provided in an embodiment of the present invention, the method comprising: Step S11: Preprocess the target infrared thermal image to obtain a preprocessed infrared thermal image; the target infrared thermal image contains at least one object to be measured; Step S12: Identify each part of the target object in the preprocessed infrared thermal image to obtain target identification information; the target object is any one of the targets in the target infrared thermal image. Step S13: Determine the central axis of the target object in the preprocessed infrared thermal image based on the target recognition information, obtain the target central axis, and generate the target region of interest in the preprocessed infrared thermal image based on the target recognition information; Step S14: Perform image analysis on the target object in the infrared thermal image based on the target's central axis and region of interest.

[0019] This embodiment provides a method for processing infrared thermal images, which fills the current technological gap of not utilizing infrared thermal imaging technology for image analysis. In this method, infrared thermal images of one or more objects to be tested are pre-acquired using a medical infrared thermal imaging device to obtain a target infrared thermal image. That is, the target infrared thermal image contains at least one object to be tested; specifically, in this application, the object to be tested refers to the human body.

[0020] Once infrared thermal images of one or more objects to be tested are acquired using medical infrared thermal imaging equipment, the target infrared thermal images can be processed according to steps S11 to S14. In practical applications, the computer program executing steps S11 to S14 can be stored in a host computer, which can then be used to perform image analysis on the target objects in the infrared thermal images. Please refer to [link to relevant documentation]. Figure 2 , Figure 2 This is a schematic diagram showing the connection between a medical infrared thermal imaging device and a host computer. Figure 2 In the diagram, 11 represents the medical infrared thermal imaging equipment, and 12 represents the host computer.

[0021] In this embodiment, to obtain an infrared thermogram that facilitates subsequent analysis, the target infrared thermogram needs to be preprocessed to obtain a preprocessed infrared thermogram. After obtaining the preprocessed infrared thermogram, various parts of the target object are identified within it to obtain target identification information. These various parts include, but are not limited to, the head, neck, shoulders, arms, legs, abdomen, and torso of the human body. The target identification information includes semantic information corresponding to each identified part of the target object. Once the target identification information is obtained, the central axis of the target object is determined in the preprocessed infrared thermogram based on this information, resulting in the target central axis. This central axis serves as an anatomical reference structure. Simultaneously, a region of interest (ROI) can be generated in the preprocessed infrared thermogram based on the target identification information. The ROI is an area automatically generated based on the target identification information and the corresponding temperature information in the preprocessed infrared thermogram, which can be used for medical measurement, analysis, annotation, or comparison. Furthermore, there can be one or multiple ROIs. For example, in practical applications, if you only want to measure the temperature of the subject's forehead, you only need to set the target region of interest to the forehead; if you want to measure the temperature of the subject's forehead, neck, shoulders, and torso, you need to set the target region of interest to the forehead, neck, shoulders, and torso.

[0022] Once the target's central axis and region of interest are obtained, image analysis operations such as medical mirror annotation, symmetry thermal distribution analysis, and rehabilitation assessment can be performed on the target object in the infrared thermal image based on the target's central axis and region of interest.

[0023] It should be noted that if the target infrared thermal image contains only one object to be measured, then it is only necessary to determine the central axis and region of interest corresponding to that object in the preprocessed infrared thermal image based on the target identification information identified in the preprocessed infrared thermal image. If the target infrared thermal image contains multiple objects to be measured, then it can be determined as follows: Figure 3 The flowchart shown is used to perform image analysis on multiple objects to be tested in an infrared thermal image.

[0024] Please see Figure 3 , Figure 3 A flowchart for image analysis of multiple objects under test in an infrared thermal image, provided as an embodiment of the present invention, includes: Step S101: Preprocess the target infrared thermal image to obtain a preprocessed infrared thermal image; Step S102: Segment N objects to be tested from the preprocessed infrared thermal image and set a unique corresponding label for each identified object to be tested; Step S103: Simultaneously identify each part of N objects to be tested, and determine the central axis and target region of interest corresponding to the N objects to be tested; Step S104: After determining the central axis and target region of interest corresponding to N test objects, output the central axis and target region of interest corresponding to each test object according to the label corresponding to each test object; Step S105: According to the labels corresponding to each test object, and in combination with the central axis and target region of interest corresponding to each test object, perform medical mirror annotation, symmetry analysis and status assessment on each test object in the target infrared thermal image.

[0025] In other words, the method provided in this application can also perform independent and parallel image analysis on multiple objects to be tested in the infrared thermal image of the target, which can be applied to application scenarios such as physical examination and rehabilitation assessment where multiple objects to be tested are detected at the same time, thus effectively improving the engineering practicality of this application in practical applications.

[0026] Clearly, this method utilizes infrared thermal imaging technology to identify various parts of the object under test and determines the central axis and region of interest based on the identified information. Since the central axis and region of interest of the object can be used for image analysis operations such as medical image annotation, symmetry analysis, and condition assessment, the method provided in this application fills the current technological gap in image analysis that does not utilize infrared thermal imaging technology.

[0027] Based on the above embodiments, this embodiment further explains and optimizes the technical solution. As a preferred implementation, the above steps: preprocessing the target infrared thermal image to obtain a preprocessed infrared thermal image, include: The target infrared thermal image is subjected to non-uniformity correction, temperature normalization, noise suppression and background attenuation processing to obtain a preprocessed infrared thermal image.

[0028] In this embodiment, during the preprocessing of the target infrared thermal image, non-uniformity correction, temperature normalization, noise suppression, and background attenuation can be performed. When performing background attenuation, a human emissivity and background temperature correction model can be introduced to effectively reduce the influence of environmental factors on the infrared thermometry results in the target infrared thermal image. This allows the target identification information and region of interest subsequently identified in the target infrared thermal image to better conform to the actual physical characteristics of medical infrared thermal imaging, thereby further enhancing the reliability of this application in medical image analysis of the target object.

[0029] Furthermore, by preprocessing the infrared thermal image of the target, the influence of environmental factors on the infrared temperature measurement results can be effectively reduced, thereby making the generated target region of interest and the extracted temperature features more consistent with the actual physical characteristics of medical infrared imaging and enhancing the medical credibility of the detection results.

[0030] Obviously, the technical solution provided in this embodiment can further increase the reliability and accuracy of this application in performing medical image analysis on the target object.

[0031] Based on the above embodiments, this embodiment further explains and optimizes the technical solution. As a preferred implementation, the above steps include: identifying various parts of the target object in the preprocessed infrared thermal image to obtain target identification information, including: The target identification information is obtained by using the attitude key point detection model and / or structural template matching method to identify various parts of the target object in the preprocessed infrared thermal image.

[0032] In this embodiment, either a posture key point detection model or a structure matching method can be used to identify the various parts of the target object in the preprocessed infrared thermal image, or a combination of the two models can be used. Alternatively, in practical applications, a human structure detection model can be used to identify the various parts of the target object in the preprocessed infrared thermal image.

[0033] As a preferred implementation, a posture key point detection model can be used to identify various parts of the target object in the preprocessed infrared thermal image.

[0034] Specifically, after obtaining the preprocessed infrared thermal image, the first step is to determine the overall contour region of the target object in the preprocessed infrared thermal image through temperature thresholding and morphological operations. Then, the main body region of the target object is determined within the overall contour region based on thermal radiation continuity and spatial connectivity. Next, under the constraint of the main body region, the various parts of the target object are identified through the posture key point detection model, such as the head center point, neck point, shoulder point, and trunk axis related points. Finally, the relative positions of the head, trunk, and limbs of the target object, as well as the boundaries of the structural regions, are determined based on the spatial topological relationship between the key structural points.

[0035] It should be noted that the posture keypoint detection model is a convolutional neural network model based on the output of infrared thermograms. Its output consists of multiple keypoint probability heatmaps. By analyzing the peak positions of the heatmaps, the coordinate values ​​corresponding to key structural points of the human body can be obtained. When using the posture keypoint detection model to identify various parts of the target object in the preprocessed infrared thermogram, data augmentation methods such as rotation, scaling, temperature perturbation, and local occlusion can be used to improve the robustness of the identification results.

[0036] Obviously, the technical solution provided in this embodiment can accurately identify the various parts of the target object in the preprocessed infrared thermal image.

[0037] Based on the above embodiments, this embodiment further explains and optimizes the technical solution. As a preferred implementation, the above steps: determining the central axis of the target object in the preprocessed infrared thermal image based on the target recognition information to obtain the target central axis, include: Based on key point fitting algorithms and / or principal direction analysis and / or statistical regression methods, and according to target identification information, the central axis of the target object is determined in the preprocessed infrared thermal image to obtain the target central axis.

[0038] In this embodiment, the central axis of the target object can be determined in the preprocessed infrared thermal image based on key point fitting algorithms, principal direction analysis, or statistical regression, according to target recognition information. Alternatively, a combination of these algorithms can be used, with target recognition information also determining the central axis of the target object in the preprocessed infrared thermal image. Of course, in practical applications, the geometric center line of the human body segmentation region can also be used, with target recognition information used to determine the central axis of the target object in the preprocessed infrared thermal image.

[0039] As a preferred implementation, the central axis of the target object can be determined in the preprocessed infrared thermal image based on the key point fitting algorithm and the target recognition information.

[0040] Please see Figure 4 , Figure 4 The flowchart of a key point fitting algorithm provided in this embodiment of the invention, which determines the central axis of the target object in a preprocessed infrared thermal image based on target recognition information, includes: Step S201: Select at least two torso structure points located in the longitudinal direction of the target object in the preprocessed infrared thermal image to obtain a set of structure points; Step S202: Based on the set of structural points, the least squares fitting method is used to fit the central axis of the target object; where the mathematical expression of the central axis of the target object is y=ax+b; a and b are the central axis parameters of the target object, y is the central axis of the target object, and x is the independent variable; Step S203: Perform time smoothing on the centerline parameters of the target object calculated from consecutive frames to obtain stable centerline parameters of the target object; Step S204: Perform medical mirror annotation, left-right symmetry thermal distribution analysis, and state assessment on the target object by using the central axis of the target object.

[0041] Although the temperature distribution on a test subject's body is influenced by factors such as metabolic activity, blood perfusion, and neural regulation, the temperature on the test subject generally exhibits a roughly symmetrical distribution on a macroscopic scale. Therefore, in practical applications, the central axis of the test subject is usually used as a symmetrical reference benchmark for its left and right structures. Based on this temperature characteristic of the test subject, its central axis can be used for medical mirroring, left-right symmetry thermal distribution analysis, and state assessment.

[0042] Obviously, the central axis of the target object can be accurately determined using the technical solution provided in this embodiment.

[0043] Based on the above embodiments, this embodiment further explains and optimizes the technical solution. As a preferred implementation, the above step of generating a target region of interest in the preprocessed infrared thermal image based on target recognition information includes: Based on structural proportion rule generation algorithms and / or temperature gradient algorithms and / or region clustering algorithms and / or rule-guided region generation algorithms, a target region of interest is generated in the preprocessed infrared thermal image according to the target recognition information.

[0044] In this embodiment, a target region of interest can be generated in the preprocessed infrared thermal image based on a structural proportion rule generation algorithm, a temperature gradient algorithm, a region clustering algorithm, or a rule-guided region generation algorithm, according to the target recognition information. Alternatively, a combination of the above-mentioned model algorithms can be used to generate the target region of interest in the preprocessed infrared thermal image based on the target recognition information.

[0045] The structural proportion rule-based generation algorithm refers to using identified structural regions as references and defining the location and range of functional regions according to the relative geometric proportions of the structure of the object under test. For example, in the head region of the target object, the forehead temperature measurement area is not arbitrarily set, but is limited to a certain height within the head region; in the shoulder and neck region, the temperature measurement area needs to be limited to a certain range between the neck and shoulder points; in the torso region, the symmetry analysis area can be generated equidistantly from the central axis of the target object. In other words, the structural proportion rule-based generation algorithm is essentially a region localization method based on the geometric relationships of the structure of the object under test.

[0046] The temperature gradient algorithm refers to the following: Since the temperature distribution of different tissue boundaries and different local areas in the infrared thermal image of the object under test at a certain moment often shows certain differences, the region of interest of the target can be generated in the preprocessed infrared thermal image based on the direction and speed of temperature change between adjacent positions in the infrared thermal image (the degree of temperature change in space).

[0047] Specifically, the temperature gradient algorithm analyzes temperature changes in the infrared thermogram of the object under test. This helps the host computer find more stable and reasonable region boundaries, or avoid areas with blurred edges, drastic temperature changes, or unsuitable measurement conditions. Alternatively, the temperature gradient algorithm can be understood as using the temperature distribution and its variation characteristics at different locations in the infrared thermogram to help determine the boundaries of the target region of interest and select stable regions. For example, within the candidate region of the forehead of the object under test, if the temperature changes too rapidly in certain areas, it may be because these areas are located at the edge of hair, occlusion, or in areas with background interference. These areas are therefore unsuitable as stable temperature measurement regions. The host computer can prioritize connected regions with more uniform temperature distribution and gentler gradient changes as the final forehead temperature measurement region.

[0048] Region clustering algorithms automatically group pixels with similar temperature values, adjacent spatial locations, and continuous distribution into the same region. Essentially, this algorithm automatically aggregates areas in an infrared thermal image that have similar temperatures and are almost connected, then combines this with the structural location of the target object to determine whether this area can be considered a region of interest. For example, within the candidate area of ​​the forehead of the target object, pixels with similar temperatures and that are interconnected can be clustered into several small regions; then, regions with suitable size, small fluctuation range, and reasonable location distribution are selected from these small regions as the forehead temperature measurement area.

[0049] The rule-guided region generation algorithm refers to the following: First, a starting point or small block is selected from the candidate regions as the starting region. Then, the algorithm gradually expands outwards from the starting region. If the pixels adjacent to the starting region meet preset conditions, the pixels around the starting region that meet the preset conditions can be aggregated into the starting region. For example, pixels with a temperature difference less than a preset value, connected to the starting region, not exceeding the boundary of the corresponding human body structure, and whose area and shape meet preset conditions can be merged into the starting region until no further expansion is possible.

[0050] As a preferred implementation, in practical applications, a region of interest (ROI) can be generated from a preprocessed infrared thermal image based on a structural proportion rule generation algorithm and target recognition information. An example is provided here to specifically illustrate the generation of a ROI from a preprocessed infrared thermal image based on a structural proportion rule generation algorithm and target recognition information.

[0051] Assuming the head structure region of the target object has been identified in the preprocessed infrared thermal image, candidate forehead regions can be determined within the identified head structure region based on ergonomic proportions. For example, the candidate forehead region might be located within the range of 0.2H to 0.4H in the vertical height of the head structure region, where H is the head height of the target object. Then, within the candidate forehead region, one or more candidate forehead temperature measurement regions are generated based on temperature gradient distribution and regional connectivity. Next, the average temperature of each candidate forehead temperature measurement region is calculated. If the average temperature of a candidate forehead temperature measurement region is less than or equal to a preset value, it can be considered a stable region of interest. If the average temperature of a candidate forehead temperature measurement region is greater than a preset threshold, it can be considered to have a high error rate and can be removed.

[0052] Of course, in practical applications, the same operation method can be used to generate temperature measurement areas in the temporal region, periorbital region, shoulder and neck region, or torso of the preprocessed infrared thermal image, and obtain a target region of interest with a relatively smaller and more refined temperature measurement area.

[0053] Furthermore, the method provided in this embodiment can automatically fit the target region of interest (ROI) into the preprocessed infrared thermogram without the need to attach any physical markers to the surface of the target object or manually designate the measurement area. This not only avoids subjective errors caused by manual operation but also improves the objectivity, accuracy, and consistency of the medical image analysis results of the target object. Moreover, in this application, when automatically fitting the ROI, the region division is not based solely on temperature extremes or simple thresholds. Instead, it combines the symmetrical structure of the human body, the constraint of the human midline, and the temperature statistical characteristics of the preprocessed infrared thermogram to generate a ROI with clear spatial semantic constraints. This not only avoids the region drift problem caused by simple hot spot detection but also improves the correspondence between the ROI and the human body structure.

[0054] Obviously, the technical solution provided in this embodiment can reliably and accurately delineate the target region of interest in the preprocessed infrared thermal image and use it for medical image analysis of the target object.

[0055] Based on the above embodiments, this embodiment further explains and optimizes the technical solution. Please refer to [link / reference]. Figure 5 , Figure 5 A flowchart illustrating another infrared thermal image processing method provided in an embodiment of the present invention. As a preferred embodiment, the above-described infrared thermal image processing method further includes: Step S301: If the number of target regions of interest is greater than 1, then calculate the temperature value corresponding to each target region of interest; Step S302: Determine the average temperature of all target regions of interest based on the temperature value corresponding to each target region of interest and the total number of all target regions of interest, and obtain the target average value; Step S303: If there is a region of interest with a temperature value greater than the average value of the target among multiple regions of interest, then mark the region of interest with a temperature value greater than the average value of the target to obtain the marked region of interest; Step S304: If the area of ​​the marked region of interest is larger than the preset area, and the duration for which the temperature value corresponding to the marked region of interest is greater than the target average value exceeds the preset duration, then the target region of interest is determined to be an abnormal region of interest, and a warning message is issued.

[0056] In practical applications, infrared thermal images of the object under test can also be used to provide early warnings of temperature anomalies in a specific area of ​​the object, allowing people to be aware of their health status in a timely manner. Specifically, if there are more than one region of interest (ROI) on the object under test, the temperature value corresponding to each ROI is first calculated. Then, based on the temperature values ​​of each ROI and the total number of ROIs, the average temperature of all ROIs is determined, resulting in the target average value.

[0057] If there is a region of interest with a temperature value greater than the average value of the target among multiple target regions of interest, it indicates that the region of interest with a temperature value greater than the average value of the target is abnormal and the target object under test may have health risks. In order to determine whether there is really a problem in the region of interest, it is necessary to mark the region of interest with a temperature value greater than the average value of the target, and obtain the marked region of interest.

[0058] If the area of ​​the marked region of interest is larger than the preset area, and the duration for which the temperature value corresponding to the marked region of interest is greater than the target average exceeds the preset duration, it indicates that there is indeed a problem with the marked region of interest, which will affect the health status of the target object. In this case, the target region of interest needs to be identified as an abnormal region of interest, and a warning message should be issued.

[0059] Obviously, the technical solution provided in this embodiment can promptly detect abnormal hot areas in the target object under test.

[0060] Based on the above embodiments, this embodiment further explains and optimizes the technical solution. Please refer to [link / reference]. Figure 6 , Figure 6 This is a flowchart illustrating another infrared thermal image processing method provided by an embodiment of the present invention. As a preferred embodiment, the above-described infrared thermal image processing method further includes: Step S401: Collect infrared thermal images of the target object according to a preset cycle to obtain multiple frames of infrared thermal images; Step S402: Process the multi-frame infrared thermal images and obtain the target region of interest corresponding to the target object in the multi-frame infrared thermal images to obtain the target set; Step S403: Determine the temperature information corresponding to all regions of interest of all targets in the target set, and sort the temperature information corresponding to all regions of interest of all targets in the target set according to the time order of the infrared thermal images to obtain time series data; Step S404: Use time series data to evaluate the state changes of the target object under test.

[0061] In practical applications, it is also possible to track the temperature information change trend of the target object in the region of interest over a period of time, and to evaluate the state changes of the target object based on the temperature information change trend of the target object, such as: evaluating the postoperative recovery of the target object, observing the inflammatory changes of the target object, or monitoring the rehabilitation process of the target object.

[0062] Specifically, firstly, infrared thermal images corresponding to the target object can be acquired according to a preset cycle to obtain multiple frames of infrared thermal images; secondly, the multiple frames of infrared thermal images are processed according to the method disclosed above to obtain the target regions of interest (ROIs) corresponding to the target object in the multiple frames of infrared thermal images, thus obtaining a target set; the number of ROIs can be one or more; then, the temperature information corresponding to all ROIs in the target set is determined, including but not limited to the average temperature value, maximum temperature value, and temperature distribution statistics of the ROIs; then, the temperature information corresponding to all ROIs in the target set is sorted according to the time sequence of the acquired infrared thermal images to obtain time series data; finally, the state changes of the target object can be evaluated based on the changes in the time series data.

[0063] Here, we will use an example to illustrate time series data in detail. Suppose that infrared thermal images of the target object are collected according to a preset period, resulting in 100 frames of infrared thermal images. Three regions of interest (ROIs) corresponding to the target object are obtained from these 100 frames of infrared thermal images: the forehead region, the abdomen region, and the neck region. Then, we calculate the average temperature value, maximum temperature value, and temperature distribution statistics corresponding to the forehead region, abdomen region, and neck region in these 100 frames of infrared thermal images.

[0064] Then, the average temperature, maximum temperature, and temperature distribution statistics of the forehead region in these 100 infrared thermal images can be sorted according to the time sequence of their acquisition, yielding the first, second, and third sequence data. Similarly, the average temperature, maximum temperature, and temperature distribution statistics of the abdomen region in these 100 infrared thermal images can be sorted according to the time sequence of their acquisition, yielding the fourth, fifth, and sixth sequence data. Furthermore, the average temperature, maximum temperature, and temperature distribution statistics of the neck region in these 100 infrared thermal images can be sorted according to the time sequence of their acquisition, yielding the seventh, eighth, and ninth sequence data.

[0065] Finally, the postoperative recovery, inflammation, or rehabilitation status of the target subjects are evaluated and analyzed based on the temperature changes of the first, second, third, fourth, fifth, sixth, seventh, eighth, and ninth sequence data.

[0066] In this application, by performing temperature statistical analysis on multiple frames of infrared thermal images and using the temperature standard deviation as a stability criterion, it is possible to automatically select stable temperature measurement areas with small temperature fluctuations, effectively suppressing instantaneous temperature measurement errors caused by breathing, minor body movements, or environmental disturbances, thereby making the assessment results of the state changes of the object under test more in line with the needs of medical testing and clinical evaluation.

[0067] Obviously, the technical solution provided in this embodiment can be used to evaluate and analyze changes in the object under test using infrared thermal imaging technology.

[0068] Please see Figure 7 , Figure 7 This is a structural diagram of an infrared thermal image processing apparatus provided in an embodiment of the present invention. The apparatus includes: Image preprocessing module 21 is used to preprocess the target infrared thermal image to obtain a preprocessed infrared thermal image; the target infrared thermal image contains at least one object to be measured; The information recognition module 22 is used to identify various parts of the target object in the preprocessed infrared thermal image to obtain target recognition information; the target object is any one of the target objects in the target infrared thermal image. The region generation module 23 is used to determine the central axis of the target object to be measured in the preprocessed infrared thermal image based on the target identification information, obtain the target central axis, and generate the target region of interest in the preprocessed infrared thermal image based on the target identification information. The image analysis module 24 is used to perform image analysis on the target object in the infrared thermal image of the target based on the target's central axis and the target's region of interest.

[0069] Preferably, the image preprocessing module 21 includes: The image preprocessing unit is used to perform non-uniformity correction, temperature normalization, noise suppression, and background attenuation processing on the target infrared thermal image to obtain the preprocessed infrared thermal image.

[0070] Preferably, the information recognition module 22 includes: The human structure of the target object in the preprocessed infrared thermal image is identified by using a posture key point detection model and / or a structural template matching method to obtain the target identification information.

[0071] Preferably, the region generation module 23 includes: The centerline generation unit is used to determine the centerline of the target object in the preprocessed infrared thermal image based on the key point fitting algorithm and / or principal direction analysis method and / or statistical regression method, and according to the target identification information, so as to obtain the target centerline.

[0072] Preferably, the region generation module 23 includes: The region generation unit is used to generate the target region of interest in the preprocessed infrared thermal image based on the target identification information, using a structure ratio rule generation algorithm and / or a temperature gradient algorithm and / or a region clustering algorithm and / or a rule-guided region generation algorithm.

[0073] Preferred options also include: The temperature calculation module is used to calculate the temperature value corresponding to each of the target regions of interest if the number of such regions is greater than one. The mean value calculation module is used to determine the mean temperature of all the target regions of interest based on the temperature value corresponding to each of the target regions of interest and the total number of all the target regions of interest, so as to obtain the target average value; The region marking module is used to mark the region of interest whose temperature value is greater than the average value of the target if there is a region of interest among the multiple regions of interest, thereby obtaining a marked region of interest; The anomaly alert module is used to determine the target region of interest as an abnormal region of interest and to issue a warning message if the area of ​​the marked region of interest is larger than a preset area and the temperature value corresponding to the marked region of interest is greater than the target average value for a duration exceeding a preset duration.

[0074] Preferred options also include: The image acquisition module is used to acquire infrared thermal images corresponding to the target object to be tested according to a preset period, so as to obtain multiple frames of infrared thermal images. The set acquisition module is used to process multiple frames of infrared thermal images and acquire the target region of interest corresponding to the target object under test in the multiple frames of infrared thermal images, so as to obtain the target set; The data sorting module is used to determine the temperature information corresponding to all the regions of interest of the targets in the target set, and sort the temperature information corresponding to all the regions of interest of the targets in the target set according to the time order of the infrared thermal images to obtain time series data; The change assessment module is used to assess the state changes of the target object under test using the time series data.

[0075] The infrared thermal image processing apparatus provided in this embodiment of the invention has the beneficial effects of the infrared thermal image processing method disclosed above.

[0076] Please see Figure 8 , Figure 8 This is a structural diagram of an electronic device provided in an embodiment of the present invention. The device includes: Memory 31 is used to store computer programs; The processor 32 is configured to implement the steps of the infrared thermal image processing method disclosed above when executing the computer program.

[0077] The electronic device provided in this embodiment of the invention has the beneficial effects of the infrared thermal image processing method disclosed above.

[0078] Accordingly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of an infrared thermal image processing method as disclosed above.

[0079] The computer-readable storage medium provided in this embodiment of the invention has the beneficial effects of the infrared thermal image processing method disclosed above.

[0080] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0081] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0082] The present invention has provided a detailed description of an infrared thermal image processing method, apparatus, device, and medium. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method of processing an infrared thermogram, characterized by, include: The target infrared thermal image is preprocessed to obtain a preprocessed infrared thermal image; The target infrared thermal image contains at least one object to be measured. The target object to be tested is identified in various parts of the preprocessed infrared thermal image to obtain target identification information; the target object to be tested is any one of the objects to be tested in the target infrared thermal image. Based on the target recognition information, the central axis of the target object to be measured is determined in the preprocessed infrared thermal image to obtain the target central axis, and the target region of interest is generated in the preprocessed infrared thermal image based on the target recognition information. Image analysis is performed on the target object in the infrared thermal image of the target based on the target's central axis and the target's region of interest.

2. The method for processing infrared thermal images according to claim 1, characterized in that, The preprocessing of the target infrared thermal image to obtain a preprocessed infrared thermal image includes: The target infrared thermal image is subjected to non-uniformity correction, temperature normalization, noise suppression, and background attenuation processing to obtain the preprocessed infrared thermal image.

3. The method for processing infrared thermal images according to claim 1, characterized in that, The step of identifying various parts of the target object in the preprocessed infrared thermal image to obtain target identification information includes: The target identification information is obtained by using a posture key point detection model and / or a structural template matching method to identify various parts of the target object in the preprocessed infrared thermal image.

4. The method for processing infrared thermal images according to claim 1, characterized in that, The step of determining the central axis of the target object in the preprocessed infrared thermal image based on the target identification information to obtain the target central axis includes: Based on the key point fitting algorithm and / or principal direction analysis method and / or statistical regression method, and according to the target identification information, the central axis of the target object to be measured is determined in the preprocessed infrared thermal image to obtain the target central axis.

5. The method for processing infrared thermal images according to claim 1, characterized in that, The step of generating a region of interest for a target in the preprocessed infrared thermal image based on the target identification information includes: The target region of interest is generated in the preprocessed infrared thermal image based on the structural proportion rule generation algorithm and / or temperature gradient algorithm and / or region clustering algorithm and / or rule-guided region generation algorithm, according to the target identification information.

6. The method for processing infrared thermal images according to claim 1, characterized in that, Also includes: If the number of the target regions of interest is greater than one, then the temperature value corresponding to each target region of interest is calculated; The average temperature of all the regions of interest is determined based on the temperature value corresponding to each of the target regions of interest and the total number of all the target regions of interest, thus obtaining the target average value; If there is a region of interest with a temperature value greater than the average value of the target among the multiple regions of interest, then the region of interest with a temperature value greater than the average value of the target is marked to obtain a marked region of interest; If the area of ​​the marked region of interest is larger than a preset area, and the duration for which the temperature value corresponding to the marked region of interest is greater than the target average value exceeds a preset duration, then the target region of interest is determined to be an abnormal region of interest, and a warning message is issued.

7. The method for processing infrared thermal images according to claim 1, characterized in that, Also includes: The infrared thermal images corresponding to the target object to be tested are collected according to a preset cycle to obtain multiple frames of infrared thermal images; The multi-frame infrared thermal images are processed, and the target region of interest corresponding to the target object under test in the multi-frame infrared thermal images is obtained to obtain the target set; The temperature information corresponding to all regions of interest of the targets in the target set is determined, and the temperature information corresponding to all regions of interest of the targets in the target set is sorted according to the time order of the infrared thermal images to obtain time series data; The time series data is used to evaluate the state changes of the target object under test.

8. An infrared thermal image processing apparatus, characterized in that, include: The image preprocessing module is used to preprocess the infrared thermal image of the target to obtain a preprocessed infrared thermal image. The target infrared thermal image contains at least one object to be measured. The information recognition module is used to identify various parts of the target object in the preprocessed infrared thermal image to obtain target recognition information; the target object is any one of the target objects in the target infrared thermal image. The region generation module is used to determine the central axis of the target object to be measured in the preprocessed infrared thermal image based on the target identification information, obtain the target central axis, and generate the target region of interest in the preprocessed infrared thermal image based on the target identification information. The image analysis module is used to perform image analysis on the target object in the infrared thermal image of the target based on the target's central axis and the target's region of interest.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the infrared thermal image processing method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of a method for processing an infrared thermal image as described in any one of claims 1 to 7.

Images