Thermal sensation evaluation method and system based on infrared thermal imaging technology

Abstract

The invention relates to the technical field of thermal sensation evaluation, in particular to a thermal sensation evaluation method and system based on an infrared thermal imaging technology, and the method comprises the steps: obtaining a target object face infrared thermal image sequence captured by an infrared thermal imager; dividing at least one pair of symmetrical physiological region pairs based on the left-right symmetrical physiological region of the human face; calculating a dynamic symmetry index of each current symmetric physiological region pair in the infrared thermal image sequence obtained in real time, comparing the dynamic symmetry index with a normal symmetry reference range in a clean environment, and generating a comprehensive symmetry breaking index SBI; a dynamic symmetry index of a facial symmetric physiological region pair is calculated and compared with a reference to generate a comprehensive symmetry breaking index SBI to quantify dust interference, the confidence coefficient of a thermal sensation evaluation result is adjusted or an evaluation strategy is switched according to the SBI level, and the interference of dust on infrared radiation transmission is recognized and quantified. The thermal sensation evaluation result is adjusted according to the interference degree, and the evaluation accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal sensation evaluation, and particularly to a thermal sensation evaluation method and system based on infrared thermal imaging technology. Background Art

[0002] In industrial production environments, such as wood processing, textile manufacturing, or metal polishing workshops, the thermal comfort of workers is crucial for work efficiency and health and safety. In the prior art, infrared thermal imaging technology is usually used to non-contact monitor the body surface temperature of workers, and combined with a thermal sensation evaluation model to guide the real-time adjustment of environmental control systems (such as ventilation, air conditioning, etc.). Specifically, an infrared thermal imager periodically captures infrared thermal radiation images of exposed parts such as the face or neck of workers, extracts temperature features (such as average temperature, standard deviation of temperature distribution), and calculates a thermal sensation index based on the correlation model between the calibrated skin temperature and subjective thermal sensation under clean air to control environmental equipment. This technology can effectively maintain a thermally comfortable environment in scenarios with high air cleanliness.

[0003] However, the above technology faces significant challenges in actual industrial applications. Due to the large amount of suspended particulate matter (such as wood chips, fibers, metal dust) generated during production activities, these dusts form a dynamically changing aerosol system in the air. When the infrared thermal imager measures, the dust particles will interact with the infrared radiation emitted by the skin: part of the radiation is absorbed or scattered, resulting in attenuation of the effective signal; at the same time, the dust itself emits interfering radiation due to temperature or friction heating, which leads to a dynamically changing deviation between the apparent temperature measured by the thermal imager and the true skin temperature of the person. The existing thermal sensation evaluation models calibrated under clean air cannot directly process this kind of disturbed temperature data, so as to avoid evaluation errors caused by dust interference and ensure that the environmental control system can be effectively adjusted based on accurate thermal sensation information; In view of the above problems, the prior art urgently needs to be improved. Summary of the Invention

[0004] The purpose of the present invention is to solve the disadvantages existing in the prior art, and to propose a thermal sensation evaluation method and system based on infrared thermal imaging technology.

[0005] In a first aspect, the present invention provides a thermal sensation evaluation method based on infrared thermal imaging technology, and the method includes the following steps: S1: Obtain a sequence of infrared thermal images of the face of a target object captured by an infrared thermal imager; S2: Divide at least one pair of symmetric physiological region pairs based on the left-right symmetric physiological regions of the human face; S3: Calculate the dynamic symmetry indexes of current symmetric physiological region pairs in the real-time acquired infrared thermal image sequence, compare the dynamic symmetry indexes with the normal symmetry benchmark range in a clean environment, and generate a comprehensive symmetry breaking index SBI, where the SBI is used to quantify the interference degree of dust on infrared radiation transmission; S4: Divide the comprehensive symmetry breaking index SBI into several levels, and dynamically adjust the confidence level of the thermal sensation evaluation result or switch the thermal sensation evaluation strategy according to the level of SBI.

[0006] Specifically, this method quantifies the interference degree of dust on infrared radiation transmission by analyzing the differences in the infrared thermal image features of the left and right symmetric regions of the human face, and then adjusts the thermal sensation evaluation result based on infrared thermography. In an industrial production environment, suspended dust will absorb and scatter infrared radiation, resulting in a deviation between the apparent temperature measured by the infrared thermal imager and the true skin temperature. This interference may be spatially non-uniform, breaking the inherent left-right thermal distribution symmetry of the human face. The method first acquires the facial infrared thermal image sequence of the target object (S1), which is the basic data for evaluation. Then, based on the physiological symmetry of the human face, the left and right symmetric physiological region pairs are divided (S2). In each frame of the image, the symmetry indexes (such as temperature difference, correlation, etc.) between these symmetric region pairs are calculated to obtain the dynamic symmetry indexes (S3). These real-time indexes are compared with the normal symmetry benchmark range established in a clean environment, and the deviation degree is calculated to generate the comprehensive symmetry breaking index SBI. The SBI value reflects the interference intensity of dust on infrared measurement in the current environment. The higher the SBI value, the greater the interference. Finally, the usage mode of the thermal sensation evaluation result is dynamically adjusted according to the level of SBI (S4). When the SBI is low, it is considered that the interference is small and the evaluation result based on the apparent temperature is highly reliable; when the SBI is high, it is considered that the interference is large, the confidence level of the evaluation result is reduced, or switched to other evaluation strategies that are not affected or less affected by dust, such as the evaluation model based on environmental sensor data. Thus, this method can identify and quantify dust interference, avoid using distorted temperature data for evaluation when the interference is severe, and improve the accuracy and robustness of thermal sensation evaluation in a dust environment.

[0007] Furthermore, this application also proposes that step S1 includes: S11: Periodically capture the original image containing the face through a fixedly installed infrared thermal imager; S12: Perform noise filtering on the original image and use the image registration technology based on feature point matching to align multiple frames of infrared thermal image sequences to the same spatial coordinate system to obtain the facial infrared thermal image sequence of the target object.

[0008] Furthermore, this application also proposes that step S2 includes: S21: Use a face recognition algorithm to locate the face region in the infrared thermal image sequence; S22: According to the preset anatomical landmarks or templates, automatically divide a number of left - right symmetric physiological region pairs with the facial mid - line as the axis of symmetry, including the left - right forehead regions, the left - right cheek regions, and the left - right eye - corner regions.

[0009] Furthermore, this application also proposes that step S3 includes: S31: Repeat the operation of step S2 for each frame of the obtained infrared thermal image sequence to divide the left - right symmetric physiological region pairs of the human face; S32: Calculate the symmetry metric index of each symmetric physiological region pair in each frame to obtain a real - time dynamic symmetry index; S33: Compare the dynamic symmetry index with the normal symmetry reference range in a clean environment, and calculate the weighted average or maximum value of the selected index deviating from the reference range, which is the comprehensive symmetry - breaking index SBI.

[0010] Furthermore, this application also proposes that the symmetry index includes the temperature difference, the temperature distribution correlation coefficient, and the similarity index.

[0011] Furthermore, this application also proposes that the establishment of the normal symmetry reference range in a clean environment includes: When the dust sensor is used to assist in judging that the air environment is clean, collect multiple frames of reference infrared thermal image sequences; Calculate the average temperature difference, the temperature distribution correlation coefficient, and the structural similarity index of each symmetry index in each symmetric physiological region pair; Statistical distribution range of each symmetry index to establish a normal symmetry reference.

[0012] Furthermore, this application also proposes that step S4 includes: S41: Divide into non - interference / slight interference, moderate interference, and severe interference levels according to the threshold of SBI; S42: Use SBI as an influencing factor for the confidence level of the thermal sensation evaluation result. If SBI is in the non - interference / slight interference level, it is determined that the confidence level of the current thermal sensation evaluation result calculated based on the apparent temperature is high, and as SBI increases, the confidence level of the evaluation result decreases.

[0013] Furthermore, this application also proposes that the switching thermal sensation evaluation strategy includes: Set a first threshold and a second threshold, where the first threshold is less than the second threshold; When the SBI is less than or equal to the first threshold, the confidence level of the thermal sensation evaluation result is high, and the original apparent temperature is used for thermal sensation evaluation; when the SBI is greater than the first threshold and less than or equal to the second threshold, the environmental control adjustment based on this measurement is temporarily suspended, and the previous control adjustment mode with high confidence level or a preset conservative control mode is maintained; when the SBI is greater than the second threshold, the infrared thermal imaging evaluation result is disabled, and the system switches to a robust thermal sensation evaluation model based on environmental sensor data; Output an alarm message to prompt possible measurement interference and suggest checking the environment or making a manual confirmation.

[0014] Furthermore, the present application also proposes that the environmental sensor data information includes: Obtain the data of the dust concentration sensor and the environmental temperature and humidity sensor; Input the data into a robust thermal sensation evaluation model under dust interference conditions calibrated offline to generate a corrected thermal sensation index.

[0015] In a second aspect, a thermal sensation evaluation system based on infrared thermal imaging technology is provided. The system includes A data acquisition module that acquires a sequence of infrared thermal images of the face of a target object captured by an infrared thermal imager; A symmetry analysis module that divides at least one pair of symmetric physiological region pairs based on the left-right symmetric physiological regions of the human face; An interference judgment module that calculates the dynamic symmetry index of each current symmetric physiological region pair in the sequence of infrared thermal images acquired in real time, compares the dynamic symmetry index with the normal symmetry reference range in a clean environment, and generates a comprehensive symmetry breaking index SBI, where the SBI is used to quantify the interference degree of dust on infrared radiation transmission; An evaluation adjustment module that divides the comprehensive symmetry breaking index SBI into several levels, and dynamically adjusts the confidence level of the thermal sensation evaluation result or switches the thermal sensation evaluation strategy according to the level of the SBI.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This technical solution uses the symmetry breaking of the facial infrared thermal image sequence as an indicator of dust interference, avoiding the difficulty of directly physically modeling and correcting complex and dynamically changing dust interference. It can effectively identify the degree of interference on infrared thermal imaging measurement in an industrial environment with dynamically changing and unevenly distributed industrial dust. This method can improve the reliability and accuracy of the thermal sensation evaluation results under dust interference, enabling the environmental control system to more accurately respond to the real thermal comfort needs of workers, avoiding misjudgments (such as misjudging the disturbed low temperature as cold or the high temperature as hot) and incorrect control actions (such as unnecessary refrigeration or heating) caused by measurement distortion, thereby enhancing the stability and effectiveness of the environmental control system, helping to improve the comfort of the working environment and potentially reducing energy consumption.

[0017] As can be seen from the above, a thermal sensation evaluation method and system based on infrared thermal imaging technology provided by this application calculates the dynamic symmetry index of the facial symmetric physiological region pairs, generates a comprehensive symmetry breaking index SBI by comparing with a benchmark to quantify dust interference, and adjusts the confidence level of the thermal sensation evaluation result or switches the evaluation strategy according to the SBI level, having the ability to identify and quantify the interference of dust on infrared radiation transmission, adjust the thermal sensation evaluation result according to the degree of interference, and improve the evaluation accuracy. Brief Description of the Drawings

[0018] Figure 1 It is a flowchart of a thermal sensation evaluation method based on infrared thermal imaging technology proposed by the present invention.

[0019] Figure 2 It is a structural diagram of a thermal sensation evaluation system based on infrared thermal imaging technology proposed by the present invention.

[0020] In the figure: 201, data acquisition module; 202, symmetry analysis module; 203, interference judgment module; 204, evaluation adjustment module. Detailed Embodiment

[0021] The following details the embodiments of the present invention. The examples of the embodiments are shown in the drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0022] The terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality" means two or more unless otherwise specifically defined.

[0023] Application scenarios: In industrial production environments, such as wood processing, textile manufacturing or metal polishing workshops, a large amount of suspended particulate matter (such as wood chips, fibers, and metal dust) will be generated due to production activities. These dusts form a dynamically changing aerosol system in the air. When the infrared thermal imager is measuring, dust particles will interact with the infrared radiation emitted by the skin: part of the radiation is absorbed or scattered, resulting in effective signal attenuation; at the same time, the dust itself emits interference radiation due to temperature or friction heating, which leads to a dynamic deviation between the apparent temperature measured by the thermal imager and the actual skin temperature of the person. The existing thermal sensation evaluation model based on clean air calibration cannot directly process this disturbed temperature data, which may lead to evaluation errors and affect the accurate adjustment of the environmental control system. Therefore, how to effectively identify and quantify the interference of dust on infrared radiation transmission, and adjust the thermal sensation evaluation strategy accordingly to avoid evaluation errors caused by dust interference and ensure that the environmental control system can be effectively adjusted based on accurate thermal sensation information is a technical problem that needs to be solved urgently.

[0024] like Figure 1 A thermal sensation assessment method based on infrared thermal imaging technology is shown, and the method comprises the following steps: S1: Acquire a facial infrared thermal image sequence of a target object captured by an infrared thermal imager; S2: dividing at least one pair of symmetrical physiological regions based on the left-right symmetrical physiological regions of the human face; S3: Calculate the dynamic symmetry index of each current symmetrical physiological region pair in the infrared thermal image sequence acquired in real time, compare the dynamic symmetry index with the normal symmetry reference range in a clean environment, and generate a comprehensive symmetry breaking index SBI, which is used to quantify the degree of interference of dust on infrared radiation transmission; S4: The comprehensive symmetry breaking index SBI is divided into several levels, and the confidence of the thermal sensation assessment result is dynamically adjusted or the thermal sensation assessment strategy is switched according to the level of SBI.

[0025] Among them, step S1 obtains the infrared thermal image sequence of the target object's face. Specifically, the original image containing the face is periodically captured by a fixedly installed infrared thermal imager, the noise of the original image is filtered, and the image registration technology based on feature point matching is used to align the multi-frame infrared thermal image sequence to the same spatial coordinate system to obtain the infrared thermal image sequence of the target object's face.

[0026] Step S2 divides the symmetrical physiological region pairs. Specifically, a facial recognition algorithm is used to locate the face region in the infrared thermal image sequence, and according to a preset anatomical landmark or template, the facial midline is used as the symmetry axis to automatically divide a number of left-right symmetrical physiological region pairs, including left-right forehead regions, left-right cheek regions, and left-right eye corner regions.

[0027] Step S3 calculates the comprehensive symmetry breaking index SBI. Specifically, repeat the operation of Step S2 for the infrared thermal image sequence obtained for each frame to divide the symmetric physiological region pairs of the human face, calculate the symmetry metric indicators of each symmetric physiological region pair in each frame, and obtain the real-time dynamic symmetry index. The symmetry indicators include temperature difference, temperature distribution correlation coefficient, and similarity index. Compare the dynamic symmetry index with the normal symmetry benchmark range in a clean environment, and calculate the weighted average or maximum value of the selected indicator deviating from the benchmark range, which is the comprehensive symmetry breaking index SBI. The establishment of the normal symmetry benchmark range in a clean environment includes: when the air environment is judged to be clean by the dust sensor, collect multiple frames of reference infrared thermal image sequences; calculate the average temperature difference, temperature distribution correlation coefficient, and structural similarity index of each symmetry indicator in each symmetric physiological region pair; statistically analyze the distribution range of each symmetry indicator, and establish the normal symmetry benchmark.

[0028] Step S4 adjusts the thermal sensation evaluation result. Specifically, it is divided into non-interference / slight interference, moderate interference, and severe interference levels according to the value of SBI. SBI is used as an influencing factor for the confidence of the thermal sensation evaluation result. If SBI is in the non-interference / slight interference level, it is determined that the confidence of the current thermal sensation evaluation result calculated based on the apparent temperature is high. As SBI increases, the confidence of the evaluation result decreases. The switching of the thermal sensation evaluation strategy includes: setting a first threshold and a second threshold, where the first threshold is less than the second threshold; when SBI is less than or equal to the first threshold, the confidence of the thermal sensation evaluation result is high, and the original apparent temperature is used for the thermal sensation evaluation; when SBI is greater than the first threshold and less than or equal to the second threshold, temporarily suspend the environmental control adjustment based on this measurement, maintain the previous control adjustment mode with high confidence or adopt a preset conservative control mode; when SBI is greater than the second threshold, disable the infrared thermal imaging evaluation result and switch to a robust thermal sensation evaluation model based on environmental sensor data; output an alarm message to prompt possible measurement interference and suggest checking the environment or making a manual confirmation. The environmental sensor data information includes: obtaining the data of the dust concentration sensor or the environmental temperature and humidity sensor; inputting the data into the robust thermal sensation evaluation model under dust interference conditions calibrated offline to generate a corrected thermal sensation index.

[0029] Specifically, this method quantifies the interference degree of dust on infrared radiation transmission by analyzing the differences in infrared thermal image features of the left and right symmetric regions of the human face, and then adjusts the thermal sensation evaluation results based on infrared thermography. In an industrial production environment, suspended dust will absorb and scatter infrared radiation, resulting in a deviation between the apparent temperature measured by the infrared thermal imager and the true skin temperature. This interference may be spatially non-uniform, breaking the inherent left-right thermal distribution symmetry of the human face. The method first obtains a sequence of facial infrared thermal images of the target object (S1), which is the basic data for evaluation. Then, based on the physiological symmetry of the human face, pairs of left and right symmetric physiological regions are divided (S2). In each frame of the image, symmetry indexes (such as temperature difference, correlation, etc.) between these pairs of symmetric regions are calculated to obtain dynamic symmetry indexes (S3). These real-time indexes are compared with the normal symmetry benchmark range established in a clean environment, and the deviation degree is calculated to generate a comprehensive symmetry breaking index SBI. The SBI value reflects the interference intensity of dust on infrared measurement in the current environment. The higher the SBI value, the greater the interference. Finally, the usage mode of the thermal sensation evaluation results is dynamically adjusted according to the level of SBI (S4). When the SBI is low, it is considered that the interference is small and the evaluation results based on the apparent temperature are highly reliable; when the SBI is high, it is considered that the interference is large, the confidence of the evaluation results is reduced, or other evaluation strategies that are not affected by dust or are less affected are switched, such as an evaluation model based on environmental sensor data. Thus, this method can identify and quantify dust interference, avoid using distorted temperature data for evaluation when the interference is severe, and improve the accuracy and robustness of thermal sensation evaluation in a dust environment.

[0030] In some specific embodiments, the system is configured with a fixedly installed infrared thermal imager to periodically capture (e.g., once per second) infrared thermal images containing the face of a worker. The images are filtered by median filtering to remove noise, and a feature point matching-based image registration algorithm is used to align consecutive frames. A face recognition algorithm locates the face region and automatically divides it into three pairs of symmetric regions, namely the left and right foreheads, the left and right cheeks, and the left and right eye corners, based on a preset template and the midline of the face. For each frame of the image, the average temperature difference of each pair of symmetric regions and the Pearson correlation coefficient of the pixel temperature distribution are calculated. These values are compared with a reference range established from data collected in a clean workshop environment. For example, the reference range may be set such that the temperature difference on the forehead is less than 0.3 °C and the correlation coefficient is greater than 0.95. The degree of deviation of the current temperature difference and correlation coefficient from the reference range is calculated, and a weighted sum is obtained to get the SBI. For example, when the SBI is less than or equal to 0.5, it is determined as a slight interference; when it is between 0.5 and 1.5 (including 1.5), it is determined as a moderate interference; and when it is greater than 1.5, it is determined as a severe interference. When the SBI is less than or equal to 0.5, the apparent temperature measured by the infrared thermal imager is input into the standard PMV model for thermal sensation evaluation. When the SBI is between 0.5 and 1.5 (including 1.5), the environmental control adjustment based on this evaluation is suspended, and the control state when the previous SBI was less than 0.5 is maintained. When the SBI is greater than 1.5, the infrared thermal imaging evaluation result is disabled, and the data of the temperature and humidity sensors and dust concentration sensors in the workshop are switched to be used. An evaluation is performed by inputting a pre-calibrated thermal sensation model considering the influence of dust, and an alarm message is output to prompt the operator.

[0031] The present application further proposes a method for obtaining a sequence of infrared thermal images of a target object's face, which includes the following steps: S11: Periodically capture an original image containing the face through a fixedly installed infrared thermal imager; S12: Remove noise from the original image and use feature point matching-based image registration technology to align a sequence of multiple infrared thermal images to the same spatial coordinate system, obtaining a sequence of infrared thermal images of the target object's face.

[0032] Among them, in step S11, by fixedly installing the infrared thermal imager at a specific position, such as above or to the side of the working area, the influence of the camera's own movement on the image stability is reduced. Periodic capture means capturing a certain number of frames per second at a preset time interval to obtain a continuous image data stream. Capturing the original image containing the face determines the target area required for subsequent processing.

[0033] Further, step S12 processes the captured original image. Noise filtering can employ filtering algorithms in the spatial domain or frequency domain, such as median filtering or Gaussian filtering, to remove the random noise in the image caused by environmental interference or the sensor itself, thereby improving the signal-to-noise ratio of the image. An image registration technique based on feature point matching is adopted. For example, algorithms such as SIFT, SURF, or ORB are used to extract stable feature points in the image. By matching the feature points between different frames, the geometric transformation parameters (such as translation, rotation, and scaling) between frames are calculated. Using the calculated transformation parameters, the image pixels of subsequent frames are mapped to the spatial coordinate system of the first frame or a reference frame, achieving the alignment of multiple frames of images. Thus, it is ensured that the same facial physiological region in the image sequence corresponds to the same image position at different time points.

[0034] Specifically, in an industrial production environment, due to the presence of suspended particulate matters such as dust, the original images captured by the infrared thermal imager may contain noise, and the workers may have head movements, resulting in changes in the facial position and posture between consecutive frames. These problems will affect the subsequent accurate division of symmetric physiological regions of the face and the calculation of symmetry metrics. Through step S11, the infrared thermal imager is fixedly installed and images are captured periodically, obtaining an original infrared thermal image sequence containing the workers' faces. Then, in step S12, noise filtering is performed on these original images to remove the random interference caused by dust, etc., improving the image quality. Subsequently, an image registration technique based on feature point matching is adopted to correct the inter-frame displacement and rotation caused by the workers' head movements, aligning the image sequence to a unified spatial reference. Thus, a clear and inter-frame aligned infrared thermal image sequence of the target object's face is obtained, providing reliable input data for accurately dividing symmetric physiological regions and calculating dynamic symmetry metrics in subsequent steps, and solving the technical problems of large noise in the original images and non-alignment between frames.

[0035] In some specific embodiments, a fixedly installed uncooled infrared thermal imager can be used, with a frame rate of 25 frames per second and a spatial resolution of 320x240 pixels. This thermal imager periodically captures infrared images containing the workers' faces. The captured original images are first subjected to noise filtering through a 3x3 median filter. Subsequently, an image registration algorithm based on ORB feature point extraction and matching is applied to the filtered image sequence. The first frame of the sequence is selected as the reference frame, ORB feature points between each subsequent frame and the reference frame are extracted, an affine transformation matrix is estimated through the RANSAC algorithm, and geometric transformation is performed on the subsequent frames using this matrix to align them to the spatial coordinate system of the reference frame. Thus, an inter-frame aligned and noise-reduced infrared thermal image sequence of the target object's face is obtained for subsequent symmetry analysis.

[0036] This application further proposes that step S2 includes: S21: Use a face recognition algorithm to locate the face region in the infrared thermal image sequence; S22: According to the preset anatomical landmarks or templates, automatically divide a number of left - right symmetric physiological region pairs with the facial mid - line as the axis of symmetry, including the left and right forehead regions, the left and right cheek regions, and the left and right eye - corner regions.

[0037] Among them, in step S21, a face recognition algorithm is used, and its function is to accurately locate the face region in the infrared thermal image sequence. Thus, subsequent processing can focus on the target region. In step S22, according to the preset anatomical landmarks or templates, with the facial mid - line as the axis of symmetry, a number of left - right symmetric physiological region pairs are automatically divided. The preset anatomical landmarks or templates provide a reference for the facial structure. The facial mid - line serves as the axis of symmetry to guide the division process, ensuring that the divided region pairs have physiological symmetry. The divided region pairs include the left and right forehead regions, the left and right cheek regions, and the left and right eye - corner regions, which are the parts of the face with left - right symmetry.

[0038] Specifically, it is necessary to accurately identify the facial region and divide it into symmetric physiological region pairs to support the calculation of subsequent symmetry metrics. Use a face recognition algorithm to locate the face region in the infrared thermal image sequence. This step finds the face boundary in the infrared thermal image sequence and limits the processing range within the face region. Thus, subsequent processing is only carried out on the face. Further, according to the preset anatomical landmarks or templates, the facial mid - line is determined as the axis of symmetry. Based on this axis of symmetry and the landmarks or templates, a number of left - right symmetric physiological region pairs are automatically divided. These region pairs include the left and right forehead regions, the left and right cheek regions, and the left and right eye - corner regions. Through these steps, the face region is accurately identified and divided into specific left - right symmetric region pairs. This provides the basic data for calculating dynamic symmetry metrics.

[0039] In some specific embodiments, a deep - learning - based face detection model is used to identify the face bounding box in the infrared thermal image. Within this bounding box, a face key - point detection algorithm is used to locate facial feature points, such as the inner eye - corners, outer eye - corners, and the tip of the nose. The facial mid - line is estimated as the vertical line connecting the mid - points of the inner eye - corners and the tip of the nose. The left and right forehead regions are defined as the regions above the left / right eyebrows and outside the facial mid - line. The left and right cheek regions are defined as the regions below the left / right eyes and outside the facial mid - line. The left and right eye - corner regions are defined as the small regions around the left / right eye - corners. These regions are automatically segmented based on the detected key - points and the calculated facial mid - line.

[0040] This application further proposes that step S3 includes: S31: Repeat the operation of step S2 for each frame of the obtained infrared thermal image sequence to divide the symmetric physiological region pairs of the human face; S32: calculating the symmetry index of each pair of symmetrical physiological regions in each frame to obtain a real-time dynamic symmetry index; S33: Compare the dynamic symmetry index with the normal symmetry reference range under a clean environment, and calculate the weighted average or maximum value of the selected index's deviation from the reference range, which is the comprehensive symmetry breaking index SBI.

[0041] Among them, for each frame of the acquired infrared thermal image sequence, a facial recognition algorithm is executed to locate the face area. According to the preset anatomical landmarks or templates, with the midline of the face as the symmetry axis, several left-right symmetrical physiological area pairs are automatically divided, such as left-right forehead areas, left-right cheek areas, and left-right eye corner areas. For each symmetrical physiological area pair divided in each frame of the image, a metric reflecting its symmetry is calculated. These indicators may include the average temperature difference between the left and right areas, the correlation coefficient of the temperature distribution, and the structural similarity index. The symmetry indicators of each symmetrical physiological area pair calculated in each frame are combined to form a dynamic symmetry index reflecting the facial symmetry state at the current moment. These dynamic symmetry indicators are compared with the normal symmetry reference range established in a clean environment in advance. The degree to which the current indicator value deviates from the reference range is calculated. The deviation degree of multiple symmetrical physiological area pairs and the deviation information of multiple frames of images can be integrated into a single value, namely, the comprehensive symmetry breaking index SBI, by means of a weighted average or maximum value. The weighted average can set weights according to the sensitivity of different areas or different indicators to interference. The maximum value directly reflects the most serious symmetry breaking situation.

[0042] Specifically, the technical solution solves the problem of how to accurately calculate the index reflecting the degree of interference from a dynamically changing environment by processing a sequence of continuously acquired infrared thermal images. First, each frame of the sequence is independently divided into facial regions and paired with symmetric regions to ensure that subsequent calculations are based on accurate region definitions. Then, the symmetry metrics of each symmetric region pair, such as temperature difference, correlation coefficient, or similarity index, are calculated frame by frame to obtain dynamic data reflecting the symmetry state at each moment. These frame-by-frame data capture the change of interference over time. Finally, these dynamic symmetry indicators are compared with the normal benchmark range established in an interference-free environment to quantify their degree of deviation. By calculating the weighted average or maximum value of these deviations, a comprehensive symmetry breaking index (SBI) is generated. SBI integrates the symmetry information of multiple frames and multiple regions into a single value, which directly quantifies the degree of interference caused by environmental factors such as dust on infrared radiation transmission. Through these steps, the scheme can accurately and real-timely evaluate the interference level, provide a reliable basis for subsequent adjustment of the confidence of thermal sensation assessment results or switching assessment strategies, and avoid assessment errors caused by interference.

[0043] In some specific embodiments, an infrared thermal image sequence containing 100 frames is obtained. For each frame in the sequence, first, a face detection algorithm is executed to locate the face bounding box. Then, within the face bounding box, according to a preset template and the facial midline, the left and right cheek regions and the left and right forehead regions are automatically divided. For each frame image, the average temperature difference and the temperature distribution correlation coefficient of the left and right cheek regions, as well as the average temperature difference and the temperature distribution correlation coefficient of the left and right forehead regions, are calculated. These calculated values are compared with the benchmark range established under a clean environment. For example, under a clean environment, the cheek temperature difference is less than 0.5°C and the correlation coefficient is greater than 0.95. The deviation value of the index exceeding the benchmark range in each frame is calculated. For example, if the cheek temperature difference in a certain frame is 1.0°C, the deviation value is 1.0 - 0.5 = 0.5°C. The weighted average of the deviation values of all regions and all indexes in these 100 frame images is calculated, or directly the maximum value among all deviation values is taken as the final comprehensive symmetry breaking index SBI. For example, if the maximum deviation value is 0.8, then SBI is 0.8. This SBI value is subsequently used to judge the interference level.

[0044] The present application further proposes that the symmetry indexes include the temperature difference, the temperature distribution correlation coefficient, and the similarity index.

[0045] Specifically, to solve the problem that the lack of a specific measurement method makes it impossible to comprehensively capture the influence of dust on the symmetry of infrared images, this solution uses the temperature difference, the temperature distribution correlation coefficient, and the similarity index as the measurement methods for dynamic symmetry indexes. After obtaining the infrared thermal image sequence of the target object's face, each frame image is processed. First, the pairs of left and right symmetric physiological regions of the human face are divided. Then, for each pair of symmetric physiological regions in each frame image, their temperature difference, temperature distribution correlation coefficient, and similarity index are calculated respectively. The temperature difference directly reflects the difference in the overall temperature levels of the symmetric regions, and the uneven absorption and scattering of infrared radiation by dust may cause this difference. The temperature distribution correlation coefficient evaluates the consistency of the temperature spatial distribution pattern within the symmetric regions, and the local interference caused by dust may change this pattern. The similarity index comprehensively considers the brightness, contrast, and structural information of the symmetric regions and can reflect the influence of dust on the image details and textures. These calculated temperature differences, temperature distribution correlation coefficients, and similarity indexes are used as the dynamic symmetry indexes for the pair of symmetric physiological regions in this frame image. By calculating these indexes for each pair of symmetric physiological regions in multiple frame images and comparing them with the normal symmetry benchmark range under a clean environment, the comprehensive symmetry breaking index SBI can be calculated. SBI quantifies the degree of interference of dust on infrared radiation transmission. Using these specific and multi-dimensional symmetry indexes can more accurately reflect the degree of damage of dust to the symmetry of infrared thermal images, thereby improving the accuracy of SBI and providing a reliable basis for subsequent dynamically adjusting the confidence level of the thermal sensation evaluation result or switching the evaluation strategy.

[0046] In some specific embodiments, taking the left and right cheek regions as an example, the calculation of the symmetry index is illustrated. First, the pixel temperature data of the left and right cheek regions are extracted from the current frame of the infrared thermal image. The average temperature T_left of all pixels in the left cheek region and the average temperature T_right of all pixels in the right cheek region are calculated, and the temperature difference is calculated as |T_left - T_right|. The pixel temperature data of the left cheek region are arranged into a vector V_left, and the pixel temperature data of the right cheek region are arranged into a vector V_right, and the Pearson correlation coefficient R between V_left and V_right is calculated. The image patches of the left and right cheek regions are used as inputs, and their structural similarity index SSIM is calculated. The calculated temperature difference, correlation coefficient R, and similarity index SSIM are used as the dynamic symmetry indexes of the left and right cheek regions in this frame of the image. The above calculations can be repeated for multiple pairs of symmetric physiological regions (such as the left and right foreheads, the left and right eye corners), and the indexes of different region pairs are weighted averaged or the maximum value is taken to obtain the overall dynamic symmetry index of this frame of the image. These indexes are subsequently used to calculate the comprehensive symmetry breaking index SBI. For example, a function can be set to perform a weighted sum of the temperature difference, (1 - R), and (1 - SSIM) to obtain a comprehensive value. The larger this value is, the more serious the symmetry breaking is.

[0047] The present application further proposes the establishment of the normal symmetry reference range in a clean environment, including: When it is determined by the dust sensor that the air environment is clean, a multi-frame reference infrared thermal image sequence is collected; The average temperature difference, temperature distribution correlation coefficient, and structural similarity index of each symmetry index in each pair of symmetric physiological regions are calculated; The distribution ranges of each symmetry index are statistically analyzed to establish the normal symmetry reference.

[0048] Among them, the establishment of the normal symmetry reference range in a clean environment provides a specific method. This method first uses the dust sensor to judge whether the current air environment meets the clean conditions to ensure that the collected data is not interfered by dust, which is a prerequisite for establishing a reliable reference. After confirming that the environment is clean, a multi-frame reference infrared thermal image sequence is collected. The multi-frame data can reflect the normal fluctuation range of the temperature and distribution of the symmetric regions on the human face in a clean environment. Then, for the collected reference image sequence, various symmetry indexes are calculated, including the average temperature difference, temperature distribution correlation coefficient, and structural similarity index. These indexes quantify the similarity degree of the left and right symmetric regions in a clean environment. Finally, statistical analysis is performed on these calculated symmetry indexes to determine their distribution ranges in a clean environment, thereby establishing the normal symmetry reference.

[0049] Specifically, in an industrial production environment, the thermal comfort of workers is particularly important for work efficiency and health and safety. When the dust sensor assists in determining that the air environment is clean, this technical solution collects multiple frames of reference infrared thermal image sequences. Calculate the average temperature difference, temperature distribution correlation coefficient, and structural similarity index of each symmetry index for each pair of symmetric physiological regions. Statistically analyze the distribution range of each symmetry index and establish a normal symmetry benchmark. This benchmark range provides a reliable reference for the comparison between the dynamic symmetry index and the benchmark, enabling the generated comprehensive symmetry breaking index SBI to accurately reflect the degree of interference of dust on infrared radiation transmission, and further supporting the adjustment of the confidence level of subsequent thermal sensation evaluation results or the switching of evaluation strategies.

[0050] In some specific embodiments, the establishment of the normal symmetry benchmark range in a clean environment can be achieved as follows: Set a dust concentration threshold, such as 100 micrograms per cubic meter. When the dust concentration in the air detected by the dust sensor is continuously lower than this threshold for a certain period of time (such as 5 minutes), it is determined that the environment is clean. At this time, start the infrared thermal imager and collect a sequence of infrared thermal images of the target object's face at a frequency of 1 frame per second, continuously collecting, for example, 60 frames of images. For each frame of the collected images, identify the facial area and divide it into pairs of symmetric physiological regions such as the left and right forehead regions, the left and right cheek regions, and the left and right eye corner regions. Calculate the average temperature difference, temperature distribution correlation coefficient, and structural similarity index for each pair of symmetric regions in each frame. For example, for the left and right forehead regions, calculate the difference between the average temperature of the left forehead and the average temperature of the right forehead; calculate the correlation coefficient between the pixel temperature distribution of the left forehead region and the pixel temperature distribution of the right forehead region; calculate the structural similarity index of the image blocks of the left and right forehead regions. Statistically analyze all the average temperature differences of the left and right forehead regions calculated from these 60 frames of images to determine their distribution range, such as calculating the average value and standard deviation, and taking the average value plus or minus twice the standard deviation as the normal benchmark range. Similar statistics are also performed on the average temperature difference, temperature distribution correlation coefficient, and structural similarity index of the left and right cheek regions and the left and right eye corner regions, respectively, to establish their respective normal benchmark ranges. Thus, the normal symmetry benchmark range in a clean environment is established, providing a reference for the comparison of symmetry indices in subsequent real-time monitoring.

[0051] This application further proposes to dynamically adjust the confidence level of the thermal sensation evaluation result or switch the thermal sensation evaluation strategy according to the level of the comprehensive symmetry breaking index SBI. Among them, step S4 includes: S41: Divide according to the threshold of SBI into levels of no interference / slight interference, moderate interference, and severe interference; S42: Take SBI as an influencing factor for the confidence level of the thermal sensation evaluation result. If SBI is in the level of no interference / slight interference, it is determined that the confidence level of the current thermal sensation evaluation result calculated based on the apparent temperature is high, and as SBI increases, the confidence level of the evaluation result decreases.

[0052] Specifically, this technical solution details the specific method for adjusting the confidence level of the thermal sensation evaluation result according to the comprehensive symmetry breaking index SBI. In step S41, by setting the threshold of SBI, the value range of SBI is divided into different levels, and these levels represent different degrees of interference of dust on infrared radiation transmission, such as no interference / slight interference, moderate interference, and severe interference. This division provides a framework for quantifying the degree of interference. In step S42, the confidence level of the thermal sensation evaluation result is dynamically adjusted using the interference levels divided in S41. When SBI is in the no interference / slight interference level, it indicates that the interference of dust on infrared radiation is very small, and at this time, the thermal sensation evaluation result calculated based on the apparent temperature is considered to have a high confidence level. As the value of SBI increases, that is, the degree of interference develops from moderate to severe, the confidence level of the evaluation result decreases. By associating SBI with specific interference levels and directly mapping these levels to the confidence level of the evaluation result, this solution provides a clear mechanism to judge and quantify the reliability of the thermal sensation evaluation result under different dust interference conditions. This solves the problem of difficult to accurately judge the reliability of the evaluation result under different degrees of interference, making the subsequent environmental control decision based on the evaluation result more robust. S41 provides the basis for classifying the degree of interference, and S42 provides the rule for adjusting the confidence level according to the classification.

[0053] This application further proposes to divide the comprehensive symmetry breaking index SBI into several levels, and dynamically adjust the confidence level of the thermal sensation evaluation result or switch the thermal sensation evaluation strategy according to the level of SBI. The switching of the thermal sensation evaluation strategy includes: Setting a first threshold and a second threshold, where the first threshold is less than the second threshold; When SBI is less than or equal to the first threshold, the confidence level of the thermal sensation evaluation result is high, and the original apparent temperature is used for thermal sensation evaluation; when SBI is greater than the first threshold and less than or equal to the second threshold, the environmental control adjustment based on this measurement is temporarily suspended, and the previous control adjustment mode with a high confidence level is maintained or a preset conservative control mode is adopted; when SBI is greater than the second threshold, the infrared thermal imaging evaluation result is disabled, and the system switches to a robust thermal sensation evaluation model based on environmental sensor data; Output an alarm message to prompt that there may be measurement interference, and it is recommended to check the environment or perform manual confirmation.

[0054] Specifically, to address the problem of reduced accuracy in thermal sensation assessment in a dusty interference environment, this solution introduces the comprehensive symmetry breaking index SBI as an indicator to measure the degree of interference of dust on infrared radiation transmission, and dynamically adjusts or switches the thermal sensation assessment strategy based on this indicator. First, two SBI thresholds are set to divide the interference degree into three levels: low, medium, and high. When the SBI value is in the low interference level (less than or equal to the first threshold), it indicates that the infrared thermal imaging data is less affected by dust and the reliability of the assessment result is high. The system directly uses the thermal sensation assessment result based on the original apparent temperature to guide environmental control. When the SBI value is in the medium interference level (greater than the first threshold and less than or equal to the second threshold), it indicates that the infrared thermal imaging data is somewhat affected, and there is a risk in directly using it for control. The system pauses the environmental control adjustment based on the current measurement result and instead maintains the previous control state or adopts a preset conservative control mode, thus avoiding incorrect adjustments caused by inaccurate data. When the SBI value is in the severe interference level (greater than the second threshold), it indicates that the infrared thermal imaging data is no longer reliable. The system completely abandons the use of the infrared thermal imaging assessment result and switches to a robust thermal sensation assessment model that utilizes environmental sensor data (such as dust concentration, temperature, and humidity) for assessment, thereby ensuring an available assessment result under severe interference conditions. At the same time, when the SBI value is in the medium or severe interference level, the system outputs an alarm message to prompt the operator to check the environment or perform manual confirmation, thereby improving the safety and practicality of the system.

[0055] In some specific embodiments, the first threshold is set to 0.1 and the second threshold is set to 0.3. When the calculated comprehensive symmetry breaking index SBI is less than or equal to 0.1, the system uses the facial apparent temperature data obtained by the infrared thermal imager, combines it with the thermal sensation model calibrated in a clean environment, calculates the thermal sensation index, and adjusts the ventilation or air conditioning equipment based on this index. When the SBI is greater than 0.1 and less than or equal to 0.3, the system does not adjust the environmental control equipment according to the currently calculated thermal sensation index, but maintains the control mode determined when the previous SBI was less than or equal to 0.1. At the same time, the system displays an alarm message "Moderate dust interference, control adjustment paused" on the operation interface. When the SBI is greater than 0.3, the system ignores the infrared thermal imager data and instead reads the data from the dust concentration sensor and the environmental temperature and humidity sensors. These data are input into a robust thermal sensation assessment model calibrated at different dust concentrations to obtain a corrected thermal sensation index, and environmental control is carried out based on this index. At the same time, the system displays an alarm message "Severe dust interference, switched to backup assessment mode, please check the environment" on the operation interface. Thus, the system can dynamically adjust the assessment and control strategies according to the actual dust interference degree, improving the accuracy of thermal sensation assessment and the robustness of environmental control in the industrial environment.

[0056] The present application further proposes that the environmental sensor data information includes: Obtain the data of the dust concentration sensor and the environmental temperature and humidity sensor; Input the data into the robust thermal sensation evaluation model under the dust interference conditions calibrated offline to generate a corrected thermal sensation index.

[0057] Among them, obtaining the data of the dust concentration sensor and the environmental temperature and humidity sensor provides alternative environmental information inputs. The dust concentration data directly reflects the intensity of the environmental factors that may interfere with infrared measurement. The environmental temperature and humidity data are the basic parameters affecting human thermal sensation and are generally not affected by the optical interference of dust in the air. These data are necessary inputs for thermal sensation evaluation. Inputting the data into the robust thermal sensation evaluation model under the dust interference conditions calibrated offline is the core of achieving robust evaluation. Using a model that has been calibrated or trained in an environment with dust interference beforehand can process the data from environmental sensors more accurately and consider the possible impact of dust on human thermal sensation or the interpretation of sensor data. The robustness of the model ensures that the system can still provide a reliable thermal sensation estimate when the infrared data fails. Offline calibration ensures the accuracy of the model before actual application. Generating the corrected thermal sensation index is the output of the evaluation process. The thermal sensation index calculated based on the environmental sensor data and the robust model is an effective estimate in the current environment with dust interference. Calling it "corrected" indicates that it may be different from the standard thermal sensation index calculated in a clean environment or is a more accurate result after considering the interference factors. This index can be used to guide the adjustment of the environmental control system.

[0058] Specifically, when the infrared thermal imaging evaluation is unavailable due to interference, the system switches to the thermal sensation evaluation scheme based on environmental sensor data. The system obtains the data of the dust concentration sensor and the environmental temperature and humidity sensor. These data are input into a robust thermal sensation evaluation model that has been calibrated or trained in an environment with dust interference beforehand. The model processes the sensor data and considers the possible impact of dust on human thermal sensation or the interpretation of sensor data, thereby generating a corrected thermal sensation index. This corrected thermal sensation index is an effective estimate in the current environment with dust interference and is used to guide the adjustment of the environmental control system. Thus, the problem of being unable to conduct an effective thermal sensation evaluation when the infrared thermal imaging evaluation is interfered is solved, providing an alternative and robust evaluation method based on environmental sensor data and ensuring the continuous operation ability of the system in a complex environment.

[0059] In some specific embodiments, the system deploys an optical dust concentration sensor and an integrated temperature and humidity sensor. The dust concentration sensor outputs the dust mass concentration per unit volume (e.g., mg / m³) by measuring the scattering or attenuation of light by air. The temperature and humidity sensor outputs the ambient temperature (e.g., °C) and relative humidity (e.g., %RH). These sensor data are periodically sent to a processing unit (e.g., an industrial PC or PLC) via an industrial communication bus (e.g., Modbus RTU). A pre-trained machine learning model (e.g., a support vector regression model) runs in the processing unit, and this model takes dust concentration, ambient temperature, and relative humidity as input features. The model is obtained through offline calibration training in a controlled or actual industrial environment containing different combinations of dust concentration, temperature, and humidity by collecting sensor data and simultaneously recording the subjective thermal sensation scores of personnel or using other reference methods not affected by dust (e.g., the calculation results of the standard PMV model based on metabolic rate, clothing, and air velocity). When the infrared thermal imaging evaluation is determined to be unavailable, the system enables this model, inputs the sensor data obtained in real time into the model, and the model outputs a corrected predicted mean vote (PMV) index. This index reflects the expected thermal sensation of personnel under the current environmental conditions and takes into account the possible effects of dust. Thus, even if the infrared measurement is disturbed, the system can still obtain a reliable thermal sensation index to adjust the operating parameters of the ventilation or air conditioning system and maintain the thermal comfort of personnel.

[0060] Reference Figure 2 , this application further proposes a thermal sensation evaluation system based on infrared thermal imaging technology. The system is applied in the steps of any of the above methods. The system includes A data acquisition module 201 that acquires a sequence of infrared thermal images of the face of a target object captured by an infrared thermal imager; A symmetry analysis module 202 that divides at least one pair of symmetric physiological region pairs based on the left-right symmetric physiological regions of the human face; An interference judgment module 203 that calculates the dynamic symmetry index of each current symmetric physiological region pair in the sequence of infrared thermal images acquired in real time, compares the dynamic symmetry index with the normal symmetry reference range in a clean environment, and generates a comprehensive symmetry breaking index SBI. SBI is used to quantify the degree of interference of dust on infrared radiation transmission; An evaluation and adjustment module 204 that divides the comprehensive symmetry breaking index SBI into several levels and dynamically adjusts the confidence level of the thermal sensation evaluation result or switches the thermal sensation evaluation strategy according to the level of SBI.

[0061] Among them, the data acquisition module 201 is configured to acquire a sequence of infrared thermal images of the face of a target object captured by an infrared thermal imager. The symmetry analysis module 202 is configured to divide at least one pair of symmetric physiological region pairs based on the left-right symmetric physiological regions of the human face. The interference judgment module 203 is configured to calculate the dynamic symmetry index of each current symmetric physiological region pair in the sequence of infrared thermal images acquired in real time, compare the dynamic symmetry index with the normal symmetry reference range in a clean environment, and generate a comprehensive symmetry breaking index SBI, where SBI is used to quantify the degree of interference of dust on infrared radiation transmission. The evaluation and adjustment module 204 is configured to divide the comprehensive symmetry breaking index SBI into several levels, and dynamically adjust the confidence level of the thermal sensation evaluation result or switch the thermal sensation evaluation strategy according to the level of SBI.

[0062] In some specific embodiments, the data acquisition module 201 receives a sequence of infrared thermal image frames of 640x480 pixels from a fixedly installed infrared thermal imager at a frequency of 1 Hz. The symmetry analysis module 202 uses a pre-trained convolutional neural network model to detect the face region in the image, and further identifies the pixel sets of the left and right forehead regions and the left and right cheek regions. For each frame of image, the interference judgment module 203 calculates the average temperature difference and the Pearson correlation coefficient of the temperature distribution of the left and right forehead regions, and the average temperature difference and the Pearson correlation coefficient of the temperature distribution of the left and right cheek regions. The normal symmetry reference range in a clean environment is set as follows: the average temperature difference of the forehead is less than 0.8 °C, and the correlation coefficient is greater than 0.9; the average temperature difference of the cheek is less than 0.6 °C, and the correlation coefficient is greater than 0.92. SBI is calculated as the weighted average of the deviation of each index from the upper limit of its normal reference range. The evaluation and adjustment module 204 sets the first threshold of SBI to 0.15 and the second threshold to 0.4. When SBI is less than or equal to 0.15, the confidence level of the thermal sensation evaluation result is set to high, and a standard thermal sensation model based on infrared apparent temperature is adopted. When SBI is greater than 0.15 and less than or equal to 0.4, the confidence level of the evaluation result decreases, and the system maintains the control mode with the previous high confidence level. When SBI is greater than 0.4, the infrared thermal imaging evaluation result is disabled, and the thermal sensation evaluation model based on the dust concentration and the data of the environmental temperature and humidity sensor is switched.

[0063] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.

Claims

1. A thermal sensation evaluation method based on infrared thermal imaging technology, characterized in that, The method includes the following steps: S1: Obtain a sequence of facial infrared thermal images of a target object captured by an infrared thermal imager; S2: Divide at least one pair of symmetric physiological region pairs based on the left - right symmetric physiological regions of the human face; S3: Calculate the dynamic symmetry index of each current symmetric physiological region pair in the real - time obtained infrared thermal image sequence, compare the dynamic symmetry index with the normal symmetry reference range in a clean environment, and generate a comprehensive symmetry breaking index SBI, where the SBI is used to quantify the interference degree of dust on infrared radiation transmission; S4: Divide the comprehensive symmetry breaking index SBI into several levels, and dynamically adjust the confidence level of the thermal sensation evaluation result or switch the thermal sensation evaluation strategy according to the level of SBI.

2. The thermal sensation evaluation method based on infrared thermal imaging technology according to claim 1, characterized in that Step S1 includes: S11: Periodically capture the original images containing the face through a fixedly installed infrared thermal imager; S12: Perform noise filtering on the original images and use an image registration technique based on feature point matching to align the multi - frame infrared thermal image sequence to the same spatial coordinate system to obtain a sequence of facial infrared thermal images of the target object.

3. The thermal sensation evaluation method based on infrared thermal imaging technology according to claim 1, characterized in that, Step S2 includes: S21: Use a face recognition algorithm to locate the face region in the infrared thermal image sequence; S22: According to the preset anatomical landmarks or templates, automatically divide several left - right symmetric physiological region pairs with the facial mid - line as the axis of symmetry, including the left - right forehead regions, left - right cheek regions, and left - right eye - corner regions.

4. A thermal sensation evaluation method based on infrared thermal imaging technology according to claim 1, characterized in that, Step S3 includes: S31: Repeat the operation of step S2 for each frame of the obtained infrared thermal image sequence to divide the symmetric physiological region pairs of the human face; S32: Calculate the symmetry metric index of each symmetric physiological region pair in each frame to obtain the real - time dynamic symmetry index; S33: Compare the dynamic symmetry index with the normal symmetry reference range in a clean environment, and calculate the weighted average or maximum value of the selected index deviating from the reference range, which is the comprehensive symmetry breaking index SBI.

5. The thermal sensation evaluation method based on infrared thermal imaging technology according to claim 4, characterized in that The symmetry index includes temperature difference, temperature distribution correlation coefficient, and similarity index.

6. The thermal sensation evaluation method based on infrared thermal imaging technology according to claim 5, characterized in that, The establishment of the normal symmetry reference range in a clean environment includes: When the air environment is judged to be clean by an auxiliary dust sensor, collect multiple frames of reference infrared thermal image sequences; Calculate the average temperature difference, temperature distribution correlation coefficient, and structural similarity index of each symmetry index in each symmetric physiological region pair; Statistical distribution range of each symmetry index to establish a normal symmetry reference.

7. A thermal sensation evaluation method based on infrared thermal imaging technology according to claim 1, characterized in that Step S4 includes: S41: Divide into no - interference / slight - interference, moderate - interference, and severe - interference levels according to the threshold of SBI; S42: Use SBI as an influencing factor for the confidence level of the thermal sensation evaluation result. When SBI is in the no - interference / slight - interference level, it is determined that the confidence level of the current thermal sensation evaluation result calculated based on the apparent temperature is high, and as SBI increases, the confidence level of the evaluation result decreases.

8. The thermal sensation evaluation method based on infrared thermal imaging technology according to claim 7, characterized in that, The switching of the thermal sensation evaluation strategy includes: Set a first threshold and a second threshold, where the first threshold is less than the second threshold; When the SBI is less than or equal to the first threshold, the confidence level of the thermal sensation evaluation result is high, and the original apparent temperature is used for the thermal sensation evaluation; when the SBI is greater than the first threshold and less than or equal to the second threshold, the environmental control adjustment based on this measurement is suspended temporarily, and the control adjustment mode with high confidence in the previous time is maintained or the preset conservative control mode is adopted; when the SBI is greater than the second threshold, the infrared thermal imaging evaluation result is disabled, and the system switches to the robust thermal sensation evaluation model based on the environmental sensor data. An alarm message is output to prompt possible measurement interference and suggest checking the environment or making a manual confirmation.

9. The thermal sensation evaluation method based on infrared thermal imaging technology according to claim 8, characterized in that, The environmental sensor data information includes: Obtain the data of the dust concentration sensor and the environmental temperature and humidity sensor; Input the data into the robust thermal sensation evaluation model under the dust interference condition calibrated offline to generate a corrected thermal sensation index.

10. A thermal sensation evaluation system based on infrared thermal imaging technology, characterized in that, The system is applied to the steps of the method described in any one of the above claims 1-9. The system includes: A data acquisition module that acquires a sequence of facial infrared thermal images of a target object captured by an infrared thermal imager; A symmetry analysis module that divides at least one pair of symmetric physiological region pairs based on the left-right symmetric physiological regions of the human face; An interference judgment module that calculates the dynamic symmetry index of each current symmetric physiological region pair in the sequence of infrared thermal images acquired in real time, compares the dynamic symmetry index with the normal symmetry reference range in a clean environment, and generates a comprehensive symmetry breaking index SBI, where the SBI is used to quantify the interference degree of dust on infrared radiation transmission; An evaluation and adjustment module that divides the comprehensive symmetry breaking index SBI into several levels, and dynamically adjusts the confidence level of the thermal sensation evaluation result or switches the thermal sensation evaluation strategy according to the level of the SBI.

Images