A modular intelligent distribution box fault early warning method based on infrared thermal imaging

Abstract

The application belongs to the technical field of image processing, and particularly relates to a modular intelligent distribution box fault early warning method based on infrared thermal imaging, which comprises the following steps: collecting continuous multiple infrared thermal images; for each image, extracting a high-temperature area and calculating a thermal centroid; according to a temperature difference value of each image relative to a previous image, obtaining a temperature rise area and calculating a temperature rise centroid; obtaining a connecting line between the temperature rise centroid and the thermal centroid, and determining a probability of each image existing a fault according to a connecting line distance and a temperature deviation on the connecting line; finally, obtaining a curve of the probability changing with time, and calculating a probability change rate, and performing early warning in response to the probability change rate being greater than a maximum probability change rate within a preset historical backtracking frame number. The application overcomes the influence of heat conduction and improves the early warning accuracy of a real fault source.

Description

Technical Field

[0001] This invention relates to the field of image processing technology. More specifically, this invention relates to a fault early warning method for modular intelligent distribution boxes based on infrared thermal imaging. Background Technology

[0002] Modular intelligent distribution boxes are the core units of modern power distribution systems. They integrate key modules such as circuit breakers, contactors, and relays in a high density. During long-term high-load operation, these electrical modules may experience localized abnormal temperature rises due to contact aging, loose screws, or other reasons. This is a major precursor to equipment burnout or even fire, seriously threatening the safe operation of the power system.

[0003] Infrared thermal imaging technology, with its advantages of non-contact and large-area temperature measurement, can capture the temperature distribution inside the distribution box in real time, becoming an effective means of fault early warning. However, in complex scenarios with dense components inside the distribution box, the heat generated by a single fault source can quickly spread to adjacent modules through heat conduction and heat convection, resulting in a blurry and connected high-temperature area on the infrared thermal image rather than a clear fault hotspot. This poses a challenge to fault early warning and location.

[0004] To pinpoint the core of heat in ambiguous high-temperature regions, current technologies often employ centroid localization algorithms. These algorithms treat infrared thermal images as static two-dimensional images and perform geometric centroid analysis, completely ignoring the dynamic physical process of heat conduction. This approach fails to deduce the true fault source from the final heat distribution pattern, resulting in fault warnings and localizations that often focus on the macroscopic center after heat diffusion, deviating from the actual heat fault source. Therefore, overcoming the influence of heat conduction and establishing a computational model for heat diffusion from the source outwards to achieve early warning and localization of fault sources within ambiguous high-temperature regions is a pressing technical challenge in the field of power equipment condition monitoring. Summary of the Invention

[0005] To address the technical problem of inaccurate fault source localization caused by neglecting the dynamic physical process of heat conduction and relying solely on geometric analysis, this invention provides a modular intelligent distribution box fault early warning method based on infrared thermal imaging. The method includes: acquiring multiple consecutive frames of infrared thermal images of the distribution box area; for each frame, identifying the temperature with the most frequent occurrences as the background temperature, and calculating the average temperature of all pixels and the average background temperature as a segmentation threshold; extracting pixels belonging to high-temperature regions based on the segmentation threshold; using the average coordinates of all pixels within the high-temperature region as the thermal centroid; and determining the phase of each frame... For the temperature difference values ​​of pixels at the same position in the previous frame image, obtain the temperature rise region and the weight value of each pixel in the temperature rise region; calculate the weighted average of the coordinates of all pixels based on the weight values ​​to determine the temperature rise centroid; obtain the line connecting the temperature rise centroid and the thermal centroid, and determine the probability of a fault in each frame image based on the distance of the line and the temperature deviation of all pixels on the line from the thermal centroid; obtain the curve of the probability of a fault in each frame image changing over time, and calculate the probability change rate at each moment; issue an early warning in response to the probability change rate exceeding the maximum probability change rate within a preset number of historical backtracking frames.

[0006] This invention introduces dynamic comparative analysis of temperature rise centroid and thermodynamic centroid. The temperature rise centroid, through weighted calculation, can accurately capture the heat source with the fastest temperature rise rate. The thermodynamic centroid represents the macroscopic geometric center after heat diffusion. By analyzing the spatial differences between the two, the temperature characteristics of the connection line, and the temporal abrupt change characteristics of the failure probability, the interference of heat conduction effect is effectively overcome, the accuracy of fault early warning is improved, and accurate backtracking of the heat source is achieved.

[0007] Preferably, the step of extracting pixels belonging to the high-temperature region includes: traversing the temperature of all pixels and marking the pixels with temperatures greater than the segmentation threshold as pixels in the high-temperature region.

[0008] This invention utilizes the pulling effect of high temperature peaks on the mean to adaptively and accurately separate pixels in high-temperature regions from bimodal temperature data containing significant high-temperature points, providing reliable data for subsequent calculation of the thermodynamic centroid and avoiding interference from non-heating regions.

[0009] Preferably, obtaining the temperature rise region includes: calculating the temperature difference between each pixel in each frame of the image and the pixel at the same position in the previous frame of the image; traversing all differences and marking the pixels with differences greater than 0 as pixels in the temperature rise region.

[0010] This invention calculates the temperature difference between the current frame and the previous frame and marks pixels with a difference greater than 0 as temperature rise areas. This method can dynamically capture pixels that are heating up and effectively filter out pixels that are not in the heat conduction and diffusion area, thereby locking the current heat source diffusion range and providing effective data for calculating the temperature rise centroid.

[0011] Preferably, the step of obtaining the weight value of each pixel in the temperature rise region includes: calculating the absolute difference in temperature between each pixel and the pixel at the same position in the previous frame image, and using the ratio of the absolute difference to the maximum value among all absolute differences of pixels as the weight value.

[0012] Preferably, the temperature rise centroid satisfies the expression: In the formula, For the first The center of mass of temperature rise in the frame image; For the first Within the temperature rise area of ​​the frame image, the first Pixels in shaft and The coordinates of the axis; For the first Within the temperature rise area of ​​the frame image, the first The weight value of each pixel; For the first The index value and number of pixels within the temperature rise area of ​​the frame image.

[0013] This invention uses the temperature rise rate of each pixel to perform a weighted average of its coordinates, ensuring that the faster the temperature rise, the greater the contribution of the point to the temperature rise centroid position, thus bringing the temperature rise centroid closer to the source of the fault and avoiding the influence of pixels in the heat conduction area on the assessment of the temperature rise centroid.

[0014] Preferably, the probability of a fault in each frame of the image satisfies an expression, including: In the formula, For the first The probability that a frame image has a fault; For the first The Euclidean distance between the temperature rise centroid and the thermal centroid of a frame image; For the first On the line connecting the frames of the image, the first Temperature of each pixel; For the first Temperature of the centroid of the frame image; To determine the absolute value sign; For the first The index value and total number of pixels on the connecting lines of the frame image; It is a natural exponential function; This is the standard normalization function.

[0015] This invention calculates the probability of failure by measuring the proximity of the two centroids and the temperature deviation of the connecting line. Only when the point with the fastest temperature rise is close to the heat accumulation point and the temperature of the connecting line changes drastically is it determined to be a high-probability failure. This effectively eliminates non-faulty heat sources, such as high-load uniform heating or noise heat sources, making the evaluation results more reliable.

[0016] Preferably, obtaining the curve of the probability of each frame of image having a fault over time includes: constructing a time-series curve of the fault probability based on the probability of each frame of image having a fault, wherein the horizontal axis is the acquisition time of each frame of image and the vertical axis is the probability of each frame of image having a fault.

[0017] Preferably, the calculation of the probability change rate at each moment includes: for each moment on the curve, calculating the first difference value of the probability of a fault existing at each moment as the probability change rate.

[0018] Preferably, the early warning includes: marking the temperature rise centroid of the frame image corresponding to the early warning time as the fault source, and generating a visual image of the fault signal.

[0019] Preferably, the step of obtaining the temperature of the pixel includes: performing grayscale processing on each frame of the image, and converting the grayscale value of the pixel in the image into a temperature value according to the thermal imager calibration parameters.

[0020] The beneficial effects of this invention are as follows:

[0021] (1) This invention comprehensively evaluates the failure probability by analyzing the spatial distance between the two centroids and the temperature deviation on the connecting line. It can effectively distinguish between the real local abnormal temperature rise and the uniform heat generation under high load. Through dual verification, it significantly reduces the false alarms caused by noise heat sources under complex working conditions.

[0022] (2) This invention does not rely on a fixed temperature threshold, but triggers an early warning by monitoring whether the rate of change of the fault probability exceeds its recent historical maximum value. This dynamic analysis method can adapt to temperature changes under different working conditions and improves the identification accuracy of fault warning. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a modular intelligent distribution box fault early warning method based on infrared thermal imaging according to the present invention. Detailed Implementation

[0024] 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, not all, of the embodiments of the present invention. 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.

[0025] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] This invention discloses a fault early warning method for modular intelligent distribution boxes based on infrared thermal imaging, referring to... Figure 1 This includes steps S1 to S5:

[0027] S1. Acquire multiple consecutive frames of infrared thermal images of the distribution box area.

[0028] It should be noted that the modular distribution box has densely packed internal components. For example, the heat generated by loose screws can be quickly conducted to adjacent modules. In order to capture this dynamic process of heat diffusion, it is necessary to continuously acquire image sequences at a fixed frame rate to provide data for subsequent analysis of the fault area.

[0029] Specifically, an infrared thermal imager is deployed in the distribution box area, with the viewing angle adjusted to cover all key modules. Infrared thermal images are continuously acquired at a preset frame rate, and the acquired images are sorted chronologically to form an image sequence. The preset frame rate is 15fps, meaning 15 frames are acquired per second; this can be adjusted by the implementers according to actual conditions. Each frame in the sequence is converted to grayscale, and the grayscale values ​​of pixels in the image are converted to temperature values ​​according to the thermal imager's calibration parameters. Simultaneously, the position of each pixel needs to be obtained. Specifically, the pixel at the bottom left corner of the image is taken as the origin, and the horizontal direction to the right from the origin is used as... The positive direction of the axis is defined by taking the vertically upward direction from the origin as... The positive direction of the axis is used to construct a Cartesian coordinate system; the position of each pixel in the image is obtained in the Cartesian coordinate system, including... Coordinates in the axial direction and Coordinates along the axis.

[0030] At this point, the temperature of each frame and each pixel has been obtained.

[0031] S2. For each frame of image, the temperature that appears most frequently is taken as the background temperature. The average temperature of all pixels and the average background temperature are calculated as the segmentation threshold. Based on the segmentation threshold, pixels belonging to the high-temperature region are extracted. The average coordinates of each pixel in the high-temperature region are taken as the thermal centroid.

[0032] It should be noted that when a fault occurs, the heat in the fault area increases significantly, appearing as a high-temperature region in the heat map. Considering the significant temperature difference between the high-temperature region and the normal background region, the temperature histogram of the entire image will show a distinct bimodal characteristic. When the high-temperature peak appears, it will cause the average temperature of the entire image to shift towards the high-temperature direction and deviate from the stable background temperature. This deviation characteristic is used to calculate the segmentation threshold, extract the high-temperature region, and avoid other non-heating regions interfering with the subsequent fault source identification accuracy.

[0033] Specifically, for each frame of the image, based on the temperature of all pixels, the number of each temperature is counted, and the temperature with the highest number of values ​​is extracted as the background temperature.

[0034] Considering that the average temperature of all pixels in the image reflects the temperature center of gravity after being raised by the high-temperature area, and is more biased towards high temperature, the segmentation threshold should be located between the background temperature representing the normal area and the temperature center of gravity raised by the high-temperature area, to ensure that the segmentation threshold can accurately segment the high-temperature area.

[0035] To identify high-temperature regions, the average temperature of all pixels and the average temperature of the background are used as the segmentation threshold. The temperature of all pixels is then iterated through, and pixels with temperatures greater than the segmentation threshold are marked as high-temperature regions.

[0036] For example, temperature data: Background temperature: 30, mean of all temperature values: 46.56, segmentation threshold: (30+46.56) / 2=38.28. The calculated segmentation threshold can adaptively separate the temperature data of the high-temperature region from the bimodal data containing significant high-temperature points.

[0037] Considering that the heat from a fault often diffuses outward from the heat center during the diffusion process, the geometric centroid of the high-temperature region can be calculated to reflect the geometric midpoint of the heat energy concentration point.

[0038] Calculate the thermal centroid, which is equal to the average coordinates of all pixels within the high-temperature region.

[0039] At this point, the thermal centroid of each frame of the image has been obtained.

[0040] S3. Based on the temperature difference between each frame of the image and the pixel at the same position in the previous frame, determine the temperature rise region and the weight value of each pixel in the temperature rise region; calculate the weighted average of the coordinates of all pixels based on the weight values ​​to obtain the temperature rise centroid.

[0041] It should be noted that the thermal centroid represents the geometric centroid of the heat accumulation area. However, under the influence of heat conduction, the rapid diffusion of heat can easily cause the thermal centroid to gradually deviate from the source point. The essential characteristic of the fault source is the generation of heat, and its temperature change is a rapid rise in a short period of time, leading all points heated by conduction. Therefore, by calculating the temperature difference between the high-temperature area in the current frame and the same position in the previous frame, all pixels in the temperature rise area are obtained. The weight of each pixel is obtained by combining the heat conduction characteristics, ensuring that the temperature rise centroid can be biased towards those areas with the fastest temperature rise, thereby enhancing the accuracy of subsequent fault point identification.

[0042] Specifically, the temperature rise area of ​​each frame relative to the previous frame is obtained as follows: calculate the temperature difference between each pixel in each frame and the pixel at the same position in the previous frame; traverse all differences and mark the pixels with differences greater than 0 as pixels in the temperature rise area.

[0043] Within the temperature rise region: The weight value of each pixel is determined based on the temperature difference between each pixel and the pixel at the same location in the previous frame; the weight value satisfies the expression:

[0044]

[0045] In the formula, For the first Within the temperature rise area of ​​the frame image, the first The weight value of each pixel; For the first Within the temperature rise area of ​​the frame image, the first The absolute difference in temperature between a pixel and a pixel at the same location in the previous frame; For the first Within the temperature rise area of ​​the frame image, the index value and number of pixels; This is the function for finding the maximum value.

[0046] in, Reflecting the The first frame in the temperature rise region of the image The change in the temperature rise rate of the nth pixel relative to the maximum temperature rise rate; the closer this value is to 1, the better the temperature rise rate of the nth pixel. The faster the temperature rises at a pixel compared to the same location in the previous frame, the more likely the pixel is to heat up. The closer a pixel is to the fault source, the greater its weight needs to be, so that the calculated temperature rise centroid is more biased towards the fastest-heating regions; conversely, if the pixel is closer to the fault source, it indicates that the pixel is closer to the fault source. A pixel may belong to a region with slow heat conduction. By assigning lower weights to pixels with low heating rates, we can prevent pixels in heat conduction regions from affecting the centroid assessment.

[0047] The temperature rise centroid is obtained by weighting the coordinates of all pixels according to their weight values; the temperature rise centroid satisfies the expression:

[0048]

[0049] In the formula, For the first The center of mass of temperature rise in the frame image; For the first Within the temperature rise area of ​​the frame image, the first Pixels in shaft and The coordinates of the axis; For the first Within the temperature rise area of ​​the frame image, the first The weight value of each pixel; For the first The index value and number of pixels within the temperature rise area of ​​the frame image.

[0050] In the formula, Reflecting the All pixels within the temperature rise area of ​​the frame image The coordinates are obtained by weighting the axis coordinates with their corresponding weight values. This coordinate calculation process ensures that pixels with faster temperature rise rates are positioned correctly. The greater the contribution of the axis coordinates to the centroid position, the more the final temperature rise centroid will be biased towards the region with the fastest temperature rise, i.e., the potential source of failure; similarly, for all pixels within the temperature rise region... The coordinates of the axes are processed in the same way.

[0051] For example, the first Temperature data within the frame image: The source of the fault is located at: , No. Temperature data within the frame image: , No. Frame image and the first The temperature difference at the same location in the frame image: Weights of pixels in the temperature rise region: ; ; Total weight of pixels in the temperature rise region: 2.056; Temperature rise centroid: (1.32, 1.49); By assigning higher weights to pixels with faster temperature rise rates, the final temperature rise centroid result is closer to the fault source.

[0052] At this point, the temperature rise centroid of each frame of the image has been obtained.

[0053] S4. Obtain the connection between the temperature rise centroid and the thermal centroid. Based on the distance of the connection and the temperature deviation between all pixels on the connection and the thermal centroid, determine the probability of a fault in each frame of the image.

[0054] It should be noted that in actual fault events, at the moment the anomaly is first detected, the heat has not been significantly dissipated. At this time, the point of concentrated heat energy and the point of fastest temperature rise should be relatively close in spatial location. However, considering that there may be non-fault interference in the acquired images, such as uniform heating of the distribution box under high load or the presence of noise heat sources, the above two points may be close in location but not necessarily a fault situation, which can easily lead to false alarms, unnecessary inspections, and waste of power resources. Therefore, by analyzing whether the temperature change along the spatial line connecting these two points conforms to the characteristics of heat conduction trend, the probability of a fault in each frame of the image is comprehensively evaluated.

[0055] Specifically, for each frame of image, the connection between the temperature rise centroid and the thermal centroid is obtained, and all pixels on the connection are extracted.

[0056] Based on the distance of the connecting lines and the temperature deviation between all pixels on the connecting lines and the thermal centroid, the probability of a fault in each image frame is determined; the probability of a fault in each image frame satisfies the expression:

[0057]

[0058] In the formula, For the first The probability that a frame image has a fault; For the first The Euclidean distance between the temperature rise centroid and the thermal centroid of a frame image; For the first On the line connecting the frames of the image, the first Temperature of each pixel; For the first Temperature of the centroid of the frame image; To determine the absolute value sign; For the first The index value and total number of pixels on the connecting lines of the frame image; It is a natural exponential function; This is the standard normalization function.

[0059] in, Reflecting the The spatial proximity of the temperature rise centroid and the thermal centroid of the frame image; the larger this value, the closer the centroid is to the temperature rise centroid and the thermal centroid is to the thermal centroid of the frame image. The closer the spatial positions of the two centroids in a frame image; Reflecting the Along the connecting lines of the frame image, the average temperature deviation of all pixels relative to the temperature rise centroid indicates that the larger the value, the more drastic the temperature change along the connecting lines, and the more the heat distribution deviates from the normal steady-state heat conduction trend. This characteristic is consistent with the thermal characteristics of a real fault. Therefore, when both the spatial proximity and average temperature deviation values ​​are large, it indicates that the first frame image... The higher the probability of a fault, the more likely there is an abnormal heat source event in the frame image; conversely, the lower the value of any indicator, the lower the probability of a faulty heat source.

[0060] At this point, the probability of a fault in each frame of the image is obtained.

[0061] S5. Obtain the curve of the probability of a fault in each frame of the image changing over time, and calculate the probability change rate at each moment; issue an early warning in response to the probability change rate being greater than the maximum probability change rate within the preset number of historical backtracking frames.

[0062] It should be noted that when a sudden fault occurs in the distribution box, such as a loose circuit breaker wiring screw, the contact resistance suddenly increases, causing a sharp rise in temperature. The probability of a fault in the corresponding frame image will be significantly different from that in a normal state frame. Therefore, by analyzing the rate of change of probability to capture the abrupt change characteristics of the fault, a real fault will cause the probability value to jump, while the change under normal operating conditions will be relatively gradual.

[0063] Specifically, the probability change rate at each moment is obtained as follows: Based on the probability of a fault in each frame of image, a time-series curve of the fault probability is constructed, where the horizontal axis represents the acquisition time of each frame of image and the vertical axis represents the probability of a fault in each frame of image; for each moment on the curve, the first-order difference value of the probability of a fault at each moment is calculated as the probability change rate at each moment.

[0064] Let the probability change rate at the current moment be... Before the current moment The maximum value of the probability change rate at each time step is In response to Greater than This indicates that the probability change rate at the current moment exceeds the upper limit of the recent change rate, which means that there may be a fault at the current moment. An early warning is issued immediately, the temperature rise centroid of the corresponding frame image at the current moment is marked as the fault source, a fault signal visualization image is generated, and relevant personnel are notified to carry out maintenance.

[0065] in, To preset the number of historical backtracking frames, the specific number of frames can be set according to the data acquisition frame rate and disturbance conditions in the actual application scenario. Its value range is [15, 75] frames. This range is set to take into account... If the value is too small, it will be overly sensitive to the maximum value of the historical probability change rate, making it susceptible to short-term fluctuations under normal operating conditions, leading to false alarms; while If the value is too large, the benchmark corresponding to the maximum historical probability change rate will be too flat, potentially including outdated instantaneous changes, which would reduce the sensitivity to newly occurring faults and lead to missed detections; therefore, in this embodiment of the invention, it is set as follows: Take 30.

Claims

1. A modular intelligent distribution box fault early warning method based on infrared thermal imaging, characterized in that, The method comprises the following steps: Collecting continuous multiple frames of infrared thermal images of a power distribution box area; For each frame of image, counting the temperature with the largest number of occurrences as the background temperature, and calculating the average value of the temperature mean value of all pixel points and the background temperature as the segmentation threshold; Based on the segmentation threshold, extracting the pixel points belonging to the high temperature area; taking the average value of the coordinates of all pixel points in the high temperature area as the thermal centroid; According to the temperature difference of the pixel points at the same position in each frame of image relative to the previous frame of image, obtaining the temperature rise area and the weight value of each pixel point in the temperature rise area; according to the weight value, performing weighted average on the coordinates of all pixel points to determine the temperature rise centroid; Obtaining the connection line between the temperature rise centroid and the thermal centroid, and determining the probability of failure of each frame of image according to the distance of the connection line and the temperature deviation of all pixel points on the connection line from the thermal centroid; The probability of failure of each frame of image satisfies: ； is the probability of failure of the frame image; is the Euclidean distance between the temperature centroid and the thermal centroid of the frame image; is the temperature of the pixel point on the line of the frame image; is the temperature of the pixel point on the line of the frame image; is the temperature of the temperature centroid of the frame image; is the index value and total number of the pixel point on the line of the frame image; is the standard normalization function; Obtaining the curve of the probability of failure of each frame of image changing with time, and calculating the probability change rate at each time; in response to the probability change rate being greater than the maximum probability change rate in the preset historical backtracking frame number, performing early warning.

2. The modular smart distribution box fault early warning method based on infrared thermal imaging according to claim 1, characterized in that, The extraction of the pixel points belonging to the high temperature area comprises: traversing the temperatures of all pixel points, and marking the pixel points with temperatures greater than the segmentation threshold as the pixel points of the high temperature area.

3. The method for modular intelligent distribution box fault early warning based on infrared thermal imaging according to claim 1, characterized in that, The acquisition of the temperature rise area comprises: calculating the difference value of the temperature of each pixel point in each frame of image from the temperature of the pixel point at the same position in the previous frame of image; traversing all difference values, and marking the pixel points with difference values greater than 0 as the temperature rise area pixel points.

4. The modular smart distribution box failure early warning method based on infrared thermal imaging of claim 1, wherein, The acquisition of the weight value of each pixel point in the temperature rise area comprises: calculating the absolute difference value of the temperature of each pixel point from the temperature of the pixel point at the same position in the previous frame of image, and taking the ratio of the absolute difference value to the maximum value of the absolute difference values of all pixel points as the weight value.

5. The method for modular intelligent distribution box fault early warning based on infrared thermal imaging according to claim 1, characterized in that, The temperature rise centroid satisfies the expression: ； In the formula, is the first frame image temperature rise center; is the first frame image temperature rise area, the first pixel point in the axis and axis coordinates; is the first frame image temperature rise area, the first pixel point weight value; is the first frame image temperature rise area, the pixel point index value and the number.

6. The modular smart distribution box failure warning method based on infrared thermal imaging according to claim 1, characterized in that, The acquisition of the curve of the probability of failure of each frame of image changing with time comprises: constructing a time sequence curve of the failure probability according to the probability of failure of each frame of image, wherein the horizontal axis is the time when each frame of image is collected, and the vertical axis is the probability of failure of each frame of image.

7. The modular smart distribution box failure warning method based on infrared thermal imaging according to claim 1, characterized in that, The calculation of the probability change rate at each time comprises: calculating the first order difference value of the probability of failure at each time as the probability change rate.

8. The modular smart distribution box failure warning method based on infrared thermal imaging of claim 1, wherein, The early warning comprises: marking the temperature rise centroid of the frame of image at the early warning time as the failure source, and generating a failure signal visualization image.

9. The modular smart distribution box failure warning method based on infrared thermal imaging of claim 1, wherein, The acquisition of the temperature of the pixel point comprises: performing grayscale processing on each frame of image, and converting the grayscale value of the pixel point in the image into a temperature value according to the calibration parameters of the thermal imager.

Images