A belt conveyor infrared monitoring and early warning method and system

Abstract

The present invention relates to the field of belt conveyor monitoring, and more specifically, the present invention relates to a belt conveyor infrared monitoring and early warning method and system. The method comprises: calculating roller abnormal values ​​and drum abnormal values ​​according to regional growth; constructing a reference window of the pixel points to be measured, and screening the seed points to be selected in the pixel points to be measured according to the reference window; performing weighted correction on the grayscale values ​​of the seed points to be selected to obtain grayscale correction values, calculating the matching rate of the seed points to be selected, and taking the seed points to be selected with a matching rate greater than a preset matching threshold as abnormal points to calculate the motor abnormal value; adding the motor abnormal value, the roller abnormal value and the drum abnormal value and normalizing them as the belt conveyor abnormal value to complete the monitoring of the belt conveyor. Through the technical solution of the present invention, the accuracy of the abnormal area identification of the belt conveyor can be improved, and the safe operation of the belt conveyor is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of belt conveyor monitoring, and more specifically, to an infrared monitoring and early warning method and system for a belt conveyor. Background Art

[0002] Belt conveyors are a type of continuous material handling equipment widely used in industrial production. Due to their continuous operation characteristics of high speed, large capacity, and long distance, they are prone to failures in key parts such as rollers, drums, and motors. Infrared thermal imaging technology measures the infrared radiation on the surface of the target object and obtains a thermal image that reflects the surface temperature distribution of the object, which can reveal the abnormal temperature rise of the object. Applying infrared thermal imaging technology to the status monitoring and fault warning of key components of belt conveyors has obvious advantages for key components of belt conveyors that have many hidden fault points and are difficult to detect directly. The temperature conditions in key areas are detected, and then the degree of abnormality is evaluated and corresponding warning information is given based on the evaluation results.

[0003] The existing Chinese patent application document with publication number CN105760883A discloses an automatic identification method for key components of belt conveyors based on infrared thermal imaging. The method uses the region growing method to segment the preprocessed image, obtain the shape feature vector, extract entropy and contrast as the texture feature vector for the target classification of key components of belt conveyors; and designs a classifier based on BP neural network to complete the classification and identification of infrared images of key components of belt conveyors using the extracted shape features and texture features.

[0004] However, since the vibration generated by the motor during operation may cause the image captured by the infrared camera to appear blurred, the above application document only uses the highest temperature point on a single image as the seed point for regional growth, which will result in low accuracy in seed point selection, thereby causing inaccurate growth areas and affecting the accuracy of the monitoring results of the belt conveyor. Summary of the invention

[0005] In order to solve the problem of inaccurate growth area, the present invention proposes an infrared monitoring and early warning method and system for a belt conveyor.

[0006] In a first aspect, the present invention discloses an infrared monitoring and early warning method for a belt conveyor, comprising: obtaining a grayscale image of a key part of a belt conveyor according to a preset sampling interval, the key parts including a motor, a roller and a drum, and calculating an abnormal value of an idler and a drum according to regional growth; taking a pixel point whose grayscale value is greater than the average grayscale value in the grayscale image of the motor as a pixel point to be tested, constructing a reference window for the pixel point to be tested, and screening a seed point to be selected from the pixel point to be tested according to the reference window; performing a weighted correction on the grayscale value of the seed point to be selected to obtain a grayscale correction value, selecting a seed point to be selected corresponding to the maximum value of the grayscale correction value as an initial seed point, calculating a matchable rate of the seed point to be selected, and taking a seed point to be selected whose matchable rate is greater than a preset matching threshold as an abnormal point to calculate a motor abnormal value; adding and normalizing the motor abnormal value, the roller abnormal value and the drum abnormal value as the belt conveyor abnormal value to complete the monitoring of the belt conveyor.

[0007] Among them, the matching rate satisfies the relationship: , Indicates The matching rate of the pixels to be tested is Represents the grayscale correction value of the initial seed point, Indicates Grayscale correction value of the pixel to be tested, represents the initial seed point and the The Euclidean distance of the pixels to be measured, Represents an exponential function.

[0008] For the motor part, by setting a reference window to screen the seed points to be selected, the pixels with inaccurate edges can be effectively identified and screened out, thereby ensuring the accuracy of the seed points to be selected, and combining the calculated matching rate to identify abnormal points. This can comprehensively monitor the operating status of the belt conveyor and promptly detect potential faults and abnormalities, thereby improving the reliability and safety of the conveyor, helping to prevent equipment failures and reduce downtime.

[0009] Preferably, the construction of a reference window for the pixel to be tested includes: for any pixel to be tested, traversing in the horizontal direction with the pixel to be tested as the center, wherein the horizontal direction includes a horizontal left direction and a horizontal right direction; drawing a grayscale frequency histogram with the grayscale value range as the horizontal coordinate and the number of pixels as the vertical coordinate, and taking the peak grayscale value closest to the red grayscale value as the marked grayscale value; taking any pixel in the horizontal direction of the pixel to be tested as the pixel to be traversed, calculating a first difference between the grayscale value of the pixel to be traversed and the grayscale value of the pixel to be tested, and calculating a second difference between the grayscale value of the pixel to be traversed and the marked grayscale value; taking the result of negative correlation mapping the sum of the first difference and the second difference as the traversal probability, and when the traversal probability is less than a preset traversal threshold, stopping the traversal to obtain the reference window for the pixel to be tested.

[0010] The traversal probability reflects the similarity of the grayscale values ​​between the pixels to be traversed and the pixels to be tested, and their consistency with the dominant grayscale trend of the image. When the traversal probability is lower than the preset threshold, it indicates that the grayscale characteristics of the pixels to be traversed and the pixels to be tested are quite different, or inconsistent with the dominant grayscale trend of the image, so the traversal is stopped.

[0011] Preferably, the construction of a reference window for the pixel to be tested also includes: for any pixel to be tested, traversing in the horizontal direction with the pixel to be tested as the center, wherein the horizontal direction includes a horizontal left direction and a horizontal right direction; drawing a coordinate system with the position of the pixel in the grayscale image as the horizontal coordinate and the grayscale value as the vertical coordinate; taking any pixel in the horizontal direction of the pixel to be tested as the pixel to be traversed, obtaining the angle between the line connecting the pixel to be traversed and the pixel to be tested and the horizontal axis, wherein the angle is not greater than a right angle; calculating the sine value of the angle, calculating the first difference between the grayscale value of the pixel to be traversed and the grayscale value of the pixel to be tested, multiplying the sine value and the normalized first difference, and when the product is greater than or equal to the preset angle threshold, stopping the traversal to obtain the reference window for the pixel to be tested.

[0012] The spatial proximity and grayscale similarity of pixels are combined to identify the local area that is most related to the pixel under test in terms of geometry and grayscale.

[0013] Preferably, the screening of the candidate seed points includes: calculating the Euclidean distance between the two endpoints in each reference window and arranging them in ascending order according to the numerical values; calculating the possibility that the pixel to be tested is the candidate seed point:

[0014] , Indicates the possibility that the pixel to be tested is the seed point to be selected. represents the first quartile, represents the third quartile, represents the interquartile range, represents the Euclidean distance between the two endpoints in the reference window, Represents a union.

[0015] when When it is equal to 1, the pixel to be tested is the seed point to be selected. When it is equal to 0, the pixel to be tested is eliminated.

[0016] The use of quartiles can effectively utilize the distribution characteristics of the data and screen out the most representative and stable seed points by excluding extreme values ​​that deviate from the main body of the data. Such a screening mechanism helps to improve the robustness of the image processing algorithm and ensure that the selected seed points can truly reflect the characteristics of the image.

[0017] Preferably, the grayscale correction value includes: taking any sampling moment as the target moment, calculating the mean grayscale value of all pixel points in the 8-neighborhood of the selected seed point at the target moment, and similarly obtaining the mean grayscale value of the selected seed point at two moments adjacent to the target moment, and taking the mean of the grayscale value means of three consecutive sampling moments as the influencing factor; normalizing the difference between the grayscale value mean of the selected seed point at the target moment and the influencing factor as a weight, and multiplying it with the grayscale value of the selected seed point to obtain the grayscale correction value.

[0018] By calculating the mean gray value of the candidate seed point at the target time and its adjacent time, the gray stability of the point in the time series is evaluated. The impact factor reflects the average performance of the gray value of the candidate seed point over time, which helps to identify stable pixels whose gray value does not change much over time.

[0019] Preferably, the calculation of the motor abnormality value includes: taking the ratio of the number of abnormal points to the number of pixels in the grayscale image as the first ratio, and taking the ratio of the mean grayscale value of all abnormal points to 255 as the second ratio; adding the first ratio and the second ratio for normalization as the motor abnormality value.

[0020] In a second aspect, the present invention discloses a belt conveyor infrared monitoring and early warning system, comprising: a processor; and a memory, wherein the memory stores computer instructions, and when the computer instructions are executed by the processor, the system executes the above-mentioned belt conveyor infrared monitoring and early warning method.

[0021] Beneficial effects of the present invention:

[0022] 1. Use the grayscale features of the image and the spatial distribution characteristics of the pixels to monitor the condition of the key components of the belt conveyor. By analyzing the grayscale values ​​of the pixels in the grayscale image and their changes, abnormal phenomena caused by component wear, heating or other failures can be captured.

[0023] 2. It can realize real-time monitoring and early warning of key parts of the conveyor, which helps to timely discover and prevent potential failures, thereby reducing downtime and improving production efficiency and safety. Through in-depth analysis of image data, this method enhances the understanding of the operating status of the conveyor and provides a scientific basis for maintenance decisions.

[0024] 3. The abnormal area of ​​the motor part obtained by using the region growing algorithm is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0026] Figure 1 The present invention is a flowchart of an infrared monitoring and early warning method for a belt conveyor according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0028] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0029] The present invention provides an infrared monitoring and early warning method for a belt conveyor. Figure 1 As shown, a belt conveyor infrared monitoring and early warning method includes steps S1 to S4, which are described in detail below.

[0030] S1, obtain the grayscale image of the key parts of the belt conveyor according to the preset sampling interval. The key parts include the motor, roller and roller, and calculate the roller anomaly value and the roller anomaly value according to the regional growth.

[0031] Infrared thermal imaging is a non-contact temperature measurement method that determines the surface temperature of an object by detecting the infrared radiation emitted by it. Therefore, infrared thermal images are selected for analysis. The original infrared image may be affected by noise, reflection and other interferences, so it is necessary to pre-process the infrared thermal image to improve its quality.

[0032] In one embodiment, an infrared thermal imager is installed at an appropriate position in the positive direction of the key parts of the belt conveyor to ensure that a complete and clear infrared thermal image of the key parts can be collected, and the key parts include the motor, rollers and drums. The infrared image is set to be collected once per second.

[0033] Crop the infrared thermal image as needed to remove unnecessary background parts, or adjust the size of the infrared thermal image to meet the needs of subsequent processing, and perform image correction, including geometric correction and radiation correction. Geometric correction can eliminate lens distortion, and radiation correction can calibrate temperature readings. Use adaptive filtering algorithms to remove noise in infrared thermal images, and use histogram equalization algorithms to enhance the contrast in infrared thermal images, making the image detail differences clearer. Grayscale the pre-processed infrared thermal images to obtain grayscale images of key parts.

[0034] It should be noted that compared with motors, the surface temperature distribution of rollers and drums is usually more uniform when they are in normal working condition. This is because rollers and drums have a single workload, good heat dissipation conditions, simple structure, single movement mode, relatively few heat sources, and good maintenance and lubrication, which makes their surface temperature distribution more uniform when they are in normal working condition.

[0035] In addition to the heat generated by mechanical friction inside the motor, there is also Joule heat generated by the current passing through the windings, and eddy current heat generated by changes in the electromagnetic field. These additional heat sources make the temperature distribution inside the motor more complicated, and may cause heat to accumulate to varying degrees in different parts of the motor.

[0036] This means that in the infrared image, the temperature difference between the overheated area and the normal area is more obvious, and the region growing algorithm can use this temperature difference to distinguish the overheated area. Therefore, the conventional region growing algorithm can be selected to obtain the overheated area of ​​the roller and drum.

[0037] For the roller grayscale image obtained at any time, the pixel corresponding to the maximum grayscale value is used as the seed point, and the remaining pixels are traversed to calculate the difference between the grayscale value of each remaining pixel and the grayscale value of the seed point. If it is less than the empirical threshold, the pixel is divided into the range of the seed point. For example, the empirical threshold is 30. Until there are no more pixels that meet the conditions to be added, the abnormal area in the roller grayscale image is obtained. The ratio of the number of pixels in the abnormal area to the number of pixels in the roller grayscale image is used as the roller abnormal value.

[0038] The abnormal value of the roller can be obtained in the same way according to the calculation method of the abnormal value of the roller.

[0039] S2, taking the pixel points whose grayscale values ​​are greater than the average grayscale value in the motor grayscale image as the pixel points to be tested, constructing a reference window for the pixel points to be tested, and screening the seed points to be selected from the pixel points to be tested according to the reference window.

[0040] Motor vibration will increase the friction between the internal parts of the motor, causing temperature changes. This temperature change will be reflected in the grayscale value of the infrared image, and the grayscale value will fluctuate abnormally. In the original image, the grayscale value of each pixel represents the surface temperature of the corresponding position on the motor surface. Motor vibration will also cause the position of the motor relative to the camera to change, causing edge blur.

[0041] In one embodiment, a grayscale frequency histogram is drawn with the grayscale value range as the horizontal axis and the number of pixels as the vertical axis, and the peak grayscale value closest to the red grayscale value is used as the marked grayscale value.

[0042] It should be noted that, according to the existing formula, the grayscale value of white is 255, the grayscale value of yellow is approximately 244.5, the grayscale value of green is approximately 150.5, the grayscale value of orange is approximately 141.5, the grayscale value of red is approximately 76.5, the grayscale value of blue is approximately 30.6, and the grayscale value of black is 0.

[0043] The pixel points whose grayscale values ​​are greater than the average grayscale value in the motor grayscale image are taken as the pixel points to be tested, and the reference window of the pixel points to be tested is constructed, including: for any pixel point to be tested, traversing along the horizontal direction with the pixel point to be tested as the center, wherein the horizontal direction includes the horizontal left direction and the horizontal right direction.

[0044] Any pixel point in the horizontal direction of the pixel point to be tested is used as the pixel point to be traversed, and the first difference between the gray value of the pixel point to be traversed and the gray value of the pixel point to be tested is calculated, and the second difference between the gray value of the pixel point to be traversed and the gray value of the mark is calculated; the sum of the first difference and the second difference is used as the traversal probability through negative correlation mapping. When the traversal probability is less than the preset traversal threshold, stop traversal and obtain the reference window of the pixel point to be tested. Negative correlation mapping can use an exponential function.

[0045] Exemplarily, the preset traversal threshold is set to 0.65, which may be specifically set by those skilled in the art.

[0046] Screening the candidate seed points according to the reference window includes: calculating the Euclidean distance between two endpoints in each reference window and arranging them in ascending order according to the numerical values.

[0047] Calculate the interquartile range, which is the difference between the third quartile and the first quartile, to measure the spread of the data.

[0048] , Indicates the possibility that the pixel to be tested is the seed point to be selected. represents the first quartile, represents the third quartile, represents the interquartile range, represents the Euclidean distance between the two endpoints in the reference window, Represents a union.

[0049] when When it is equal to 1, the pixel to be tested is the seed point to be selected. When it is equal to 0, the pixel to be tested is eliminated.

[0050] It should be noted that quartiles, also known as quartile points, refer to the values ​​at three dividing points in statistics when all values ​​are arranged from small to large and divided into four equal parts. The first quartile is also called the lower quartile, which is the minimum value of the data points in the lower 25% of the data set. The second quartile is also called the median, which is the middle value of the data. When the data set contains an odd number of data points, the median is the middle number; when the data set contains an even number of data points, the median is the average of the two middle numbers. The third quartile is also called the upper quartile, which is the minimum value of the data points in the upper 25% of the data set.

[0051] In one embodiment, constructing a reference window for the pixel to be measured further includes:

[0052] For any pixel to be tested, traverse along the horizontal direction with the pixel to be tested as the center, wherein the horizontal direction includes a horizontal left direction and a horizontal right direction.

[0053] Constructing a coordinate system, specifically including: the origin of the coordinate system is the leftmost pixel of the grayscale image in the horizontal direction where the reference window is located. Since the pixels are arranged at equal intervals in the horizontal direction, the coordinate system is drawn with the position of the pixel in the grayscale image as the horizontal coordinate and the grayscale value of the pixel as the vertical coordinate; any pixel in the horizontal direction of the pixel to be measured is used as the pixel to be traversed, and the angle between the line connecting the pixel to be traversed and the pixel to be measured and the horizontal axis is obtained, and the angle is not greater than a right angle. It should be explained that the grayscale values ​​of the pixels in the horizontal direction are different, so the line connecting the two pixels will intersect with the horizontal axis after being extended, resulting in an angle, wherein the horizontal axis is the horizontal coordinate of the coordinate system.

[0054] Calculate the sine value of the angle, calculate the first difference between the grayscale value of the pixel to be traversed and the grayscale value of the pixel to be tested, multiply the sine value and the normalized first difference, and when the product is greater than or equal to the preset angle threshold, stop traversing and obtain the reference window of the pixel to be tested. Exemplarily, the preset angle threshold is set to 0.6.

[0055] Exemplarily, the grayscale value of the pixel to be tested is 76. First, determine the horizontal left length of the reference window of the pixel to be tested. The left pixel to be traversed adjacent to the pixel to be tested is 79. Connect the pixel to be tested and the pixel to be traversed in the coordinate system and extend them to form an angle with the horizontal axis, which is less than or equal to 90 degrees. Calculate the sine value of the angle. The larger the sine value, the greater the difference between the pixel to be traversed and the pixel to be tested, and the more the pixel to be traversed deviates from the pixel to be tested. In addition, through the constraint of the first difference, the grayscale value is prevented from increasing all the time.

[0056] Similarly, the calculation method of the reference window's horizontal length to the right is the same as that to the left.

[0057] S3, performing weighted correction on the grayscale value of the selected seed point to obtain a grayscale correction value, selecting the selected seed point corresponding to the maximum value of the grayscale correction value as the initial seed point, calculating the matching rate of the selected seed point, and taking the selected seed point whose matching rate is greater than the preset matching threshold as the abnormal point to calculate the motor abnormal value.

[0058] In one embodiment, any sampling moment is taken as the target moment, and the mean grayscale value of all pixels in the neighborhood of the candidate seed point at the target moment is calculated. Similarly, the mean grayscale value of the candidate seed point at two moments adjacent to the target moment is obtained, and the mean grayscale value of three consecutive sampling moments is taken as the influencing factor; the difference between the mean grayscale value of the candidate seed point at the target moment and the influencing factor is normalized as a weight, multiplied by the grayscale value of the candidate seed point, and a grayscale correction value is obtained.

[0059] Exemplarily, the target time is set to the 1000th second, and the grayscale value mean of all pixels in the 8-neighborhood of the candidate seed point at the 1000th second is calculated, and the grayscale value mean of all pixels in the 8-neighborhood of the candidate seed point at the 999th second and the 1001st second is calculated respectively. The three grayscale value means are averaged, and the grayscale value mean at the 1000th second is normalized by the difference between the grayscale value mean and the grayscale value at the 1000th second as the weight, which is multiplied by the grayscale value of the candidate seed point to obtain the grayscale correction value.

[0060] The grayscale correction value of each candidate seed point is obtained by traversing, and the candidate seed point corresponding to the maximum grayscale correction value is selected as the initial seed point.

[0061] Calculate the matching rate of the seed point to be selected, and the matching rate satisfies the relationship:

[0062] , Indicates The matching rate of the pixels to be tested is Represents the grayscale correction value of the initial seed point, Indicates Grayscale correction value of the pixel to be tested, represents the initial seed point and the The Euclidean distance of the pixels to be measured, Represents an exponential function.

[0063] The candidate seed points whose matching rate is greater than the preset matching threshold are regarded as abnormal points. Exemplarily, the preset matching threshold is set to 0.7.

[0064] The ratio of the number of abnormal points to the number of pixels in the grayscale image is taken as the first ratio, and the ratio of the mean grayscale value of all abnormal points to 255 is taken as the second ratio; the first ratio and the second ratio are added and normalized to obtain the motor abnormal value.

[0065] S4, adding the motor abnormal value, the roller abnormal value and the drum abnormal value and normalizing them as the belt conveyor abnormal value, completing the monitoring of the belt conveyor.

[0066] In one embodiment, for any sampling moment, the motor abnormal value, the roller abnormal value and the drum abnormal value are added and normalized to be the belt conveyor abnormal value at the sampling moment.

[0067] Traverse to obtain the belt conveyor abnormal value at each sampling moment, set the abnormal threshold, and when the belt conveyor abnormal value is greater than the abnormal threshold, mark the belt conveyor abnormal value. When the marked belt conveyor abnormal value accounts for one fortieth of all belt conveyor abnormal values, generate and send an alarm.

[0068] For example, there are now 1000 belt conveyor anomaly values, 20 of which are marked. At this time, it is believed that the belt conveyor has no abnormality. Then the belt conveyor anomaly value at the first sampling moment among the 1000 belt conveyor anomaly values ​​is eliminated, and the belt conveyor anomaly value at the 1001st sampling moment and the remaining 999 belt conveyor anomaly values ​​constitute a new belt conveyor anomaly value monitoring sequence.

[0069] An embodiment of the present invention further discloses an infrared monitoring and early warning system for a belt conveyor, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an infrared monitoring and early warning method for a belt conveyor according to the present invention is implemented.

[0070] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0071] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0072] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0073] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A belt conveyor infrared monitoring and early warning method, characterized in that: include: Obtain grayscale images of key parts of the belt conveyor according to preset sampling intervals. Key parts include motors, rollers, and drums. Calculate roller anomaly values ​​and drum anomaly values ​​based on regional growth. Pixels with grayscale values ​​greater than the average grayscale value in the motor grayscale image are taken as pixels to be tested, a reference window for the pixels to be tested is constructed, and seed points to be selected are screened from the pixels to be tested according to the reference window; Performing weighted correction on the grayscale value of the seed point to be selected to obtain a grayscale correction value, selecting the seed point to be selected corresponding to the maximum value of the grayscale correction value as the initial seed point, calculating the matching rate of the seed point to be selected, and taking the seed point to be selected whose matching rate is greater than a preset matching threshold as an abnormal point to calculate the abnormal value of the motor; The abnormal value of the motor, the abnormal value of the roller and the abnormal value of the drum are added and normalized as the abnormal value of the belt conveyor to complete the monitoring of the belt conveyor; Among them, the matching rate satisfies the relationship: , Indicates The matching rate of the pixels to be tested is Represents the grayscale correction value of the initial seed point, Indicates Grayscale correction value of the pixel to be tested, represents the initial seed point and the The Euclidean distance of the pixels to be measured, represents the exponential function; The construction of the reference window of the pixel to be measured also includes: For any pixel to be tested, traverse along the horizontal direction with the pixel to be tested as the center, wherein the horizontal direction includes a horizontal left direction and a horizontal right direction; Draw a coordinate system with the position of the pixel in the grayscale image as the horizontal coordinate and the grayscale value as the vertical coordinate; Any pixel point in the horizontal direction of the pixel point to be tested is used as the pixel point to be traversed, and the angle between the line connecting the pixel point to be traversed and the pixel point to be tested and the horizontal axis is obtained, and the angle is not greater than a right angle; Calculate the sine value of the angle, calculate the first difference between the grayscale value of the pixel to be traversed and the grayscale value of the pixel to be tested, multiply the sine value and the normalized first difference, and when the product is greater than or equal to the preset angle threshold, stop traversing and obtain the reference window of the pixel to be tested; The grayscale correction value includes: Take any sampling moment as the target moment, calculate the mean gray value of all pixels in the 8-neighborhood of the candidate seed point at the target moment, and similarly obtain the mean gray value of the candidate seed point at two moments adjacent to the target moment, and take the mean of the mean gray values ​​of three consecutive sampling moments as the influencing factor; The difference between the mean grayscale value of the seed point to be selected at the target time and the influencing factor is normalized as the weight, and multiplied by the grayscale value of the seed point to be selected to obtain the grayscale correction value.

2. The infrared monitoring and early warning method for a belt conveyor according to claim 1 is characterized in that: The construction of a reference window for the pixel to be measured comprises: For any pixel to be tested, traverse along the horizontal direction with the pixel to be tested as the center, wherein the horizontal direction includes a horizontal left direction and a horizontal right direction; Draw a grayscale frequency histogram with the grayscale value range as the horizontal axis and the number of pixels as the vertical axis, and take the peak grayscale value closest to the red grayscale value as the marked grayscale value; Take any pixel point in the horizontal direction of the pixel point to be tested as the pixel point to be traversed, calculate the first difference between the gray value of the pixel point to be traversed and the gray value of the pixel point to be tested, and calculate the second difference between the gray value of the pixel point to be traversed and the gray value of the mark; The result of negative correlation mapping of the sum of the first difference and the second difference is used as the traversal probability. When the traversal probability is less than a preset traversal threshold, the traversal is stopped to obtain a reference window of the pixel to be measured.

3. The infrared monitoring and early warning method for a belt conveyor according to claim 1 is characterized in that: The screening of the seed points to be selected comprises: Calculate the Euclidean distance between the two endpoints in each reference window and sort them in ascending order according to the numerical value; Calculate the possibility that the pixel to be tested is the seed point to be selected: , Indicates the possibility that the pixel to be tested is the seed point to be selected. represents the first quartile, represents the third quartile, represents the interquartile range, represents the Euclidean distance between the two endpoints in the reference window, represents a union; when When it is equal to 1, the pixel to be tested is the seed point to be selected. When it is equal to 0, the pixel to be tested is eliminated.

4. The infrared monitoring and early warning method for a belt conveyor according to claim 1 is characterized in that: The calculating of the motor abnormal value comprises: The ratio of the number of outliers to the number of pixels in the grayscale image is taken as a first ratio, and the ratio of the mean of the grayscale values ​​of all outliers to 255 is taken as a second ratio; The first ratio and the second ratio are added and normalized to obtain a motor abnormality value.

5. A belt conveyor infrared monitoring and early warning system, characterized in that: include: Processor; and A memory storing computer instructions, wherein when the computer instructions are executed by a processor, the system executes a belt conveyor infrared monitoring and early warning method according to any one of claims 1 to 4.

Images