A fault detection method for scraper chains of scraper conveyors

Abstract

The present invention relates to the technical field of image processing, and specifically discloses a method for detecting faults in a scraper chain of a scraper conveyor, which includes the following steps: collecting images of the scraper chain at each moment in the rotating state; calculating the degree of abnormality of the scraper chain images at each moment within a preset time period, and when the degree of abnormality of the scraper chain images at all moments within the preset time period is greater than a threshold value, it is determined that the scraper chain has a fault; the degree of abnormality is inversely correlated with both the similarity of the horizontal features and the similarity of the vertical features between the scraper chain image at the current moment and the scraper chain image at a previously set moment; the horizontal feature is a row mean sequence formed by the means of the tendency degrees of the pixel points in each row of the scraper chain image, and the vertical feature is a column mean sequence formed by the means of the tendency degrees of the pixel points in each column of the scraper chain image. The solution of the present invention can accurately distinguish the scraper chain and the sprocket in the scraper chain image, and the fault detection result is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly relates to a method for detecting faults of a scraper chain for a scraper conveyor. Background Art

[0002] A scraper conveyor is an important coal mine machinery and plays an important role in the process of transporting materials in a coal mine. A scraper conveyor is a device that drives the movement of a scraper chain through the rotation of a sprocket to achieve material transportation. Therefore, the normal operation of the scraper chain is crucial for material transportation. If a fault occurs in the scraper chain, such as chain jamming, chain derailment, or even scraper chain breakage, it will not only affect material transportation and reduce transportation efficiency, but also may cause safety accidents and property losses. Therefore, it is necessary to detect faults of the scraper chain in order to timely discover the faults of the scraper chain and perform manual intervention.

[0003] In the prior art, manual methods are usually used to detect faults of the scraper chain. However, this method relies on the observation of human eyes, resulting in low accuracy and efficiency of fault detection. With the development of image processing technology, the application of image processing technology to the fault detection of the scraper chain is becoming more and more common, which can improve the accuracy and efficiency of fault detection. The steps of using image processing technology to detect faults of the scraper chain include: collecting images of the scraper chain at multiple moments; separating the scraper chain and the sprocket in the image through an image segmentation method; comparing the position information of the scraper chain at the current moment with that at the previous moment. If the position change of the scraper chain at the current moment is greater than a set value, it is determined that the scraper chain has a fault.

[0004] However, in the collected images of the scraper chain, there will be similar situations between the scraper chain and the sprocket. During image segmentation, it is difficult to accurately separate the scraper chain and the sprocket, and there will be a situation where the sprocket is misclassified into the scraper chain. Therefore, when comparing the position information of the scraper chain, due to the inaccurate image of the separated scraper chain, the comparison result is also inaccurate, resulting in low accuracy of scraper chain fault detection. Summary of the Invention

[0005] The present invention provides a method for detecting faults of a scraper chain for a scraper conveyor, aiming to solve the technical problem in the prior art that it is difficult to accurately distinguish the scraper chain and the sprocket in the scraper chain image, resulting in low accuracy of scraper chain fault detection.

[0006] The present invention provides a method for detecting faults of a scraper chain for a scraper conveyor, including the following steps:

[0007] Collect images of the scraper chain at each moment in the rotating state;

[0008] Calculate the abnormality degree of the scraper chain images at each moment within a preset time period. When the abnormality degrees of the scraper chain images at all moments within the preset time period are greater than the threshold value, it is determined that the scraper chain has a fault;

[0009] Among them, the abnormality degree is inversely correlated with both the similarity of the horizontal features and the similarity of the vertical features between the scraper chain image at the current moment and the scraper chain image at its set moment; the horizontal feature is a row mean sequence composed of the means of the inclination degrees of the pixel points in each row of the scraper chain image, and the vertical feature is a column mean sequence composed of the means of the inclination degrees of the pixel points in each column of the scraper chain image; the inclination degree is positively correlated with the value obtained by normalizing the difference between the pixel values of the corresponding pixel point in the first color channel and the second color channel, and is positively correlated with the value obtained by normalizing the difference between the pixel values of each reference pixel point in the first color channel and the second color channel; the reference pixel point is the first pixel point after a plurality of rays radiating outward from the corresponding pixel point pass through the nearest edge; the first color channel is the color channel with the largest pixel value among the pixels of the scraper chain color, and the second color channel is the color channel with the largest pixel value among the pixels of the sprocket color.

[0010] In the above solution, by calculating the inclination degree of each pixel point, the scraper chain and the sprocket in the scraper chain image can be accurately distinguished. And by comparing the similarity of the horizontal features and the similarity of the vertical features in the scraper chain image, the abnormality degree of the scraper chain image is obtained, making the calculation result more accurate.

[0011] Preferably, the abnormality degree of the scraper chain image at the th moment is :

[0012] ;

[0013] In the formula, is the similarity of the horizontal features between the scraper chain images at the th moment and the th moment, is the similarity of the vertical features between the scraper chain images at the th moment and the th moment. The th moment is the set moment, and the th moment is the moment closest to the maximum value of the similarity of the horizontal features of the scraper chain image at the th moment, is the exponential function with the natural constant as the base.

[0014] In the above solution, the similarity of the horizontal features and the similarity of the vertical features of the scraper chain image comprehensively and accurately reflect the abnormality degree of the scraper chain image. And through the exponential function, the inverse correlation relationship between the abnormality degree and the similarity can be reflected.

[0015] Preferably, the tendency degree of the th pixel point in the scraper chain image at the th moment is :

[0016] ;

[0017] In the formula, is the pixel value of the th pixel point in the scraper chain image at the th moment in the first color channel, is the pixel value of the th pixel point in the scraper chain image at the th moment in the second color channel, is the pixel value of the th reference pixel point of the th pixel point in the scraper chain image at the th moment in the first color channel, is the pixel value of the th reference pixel point of the th pixel point in the scraper chain image at the th moment in the second color channel, is the normalization function, is the adjustment parameter, , is the total number of reference pixel points of the th pixel point in the scraper chain image at the th moment, is the exponential function with the natural constant as the base.

[0018] In the above solution, the tendency degree is corrected by the difference in pixel values of each reference pixel point in the first color channel and the second color channel, preventing the calculated tendency degree from being too small due to the wear of the protective paint on the scraper chain, resulting in inaccurate calculation results.

[0019] Preferably, the similarity is calculated using cosine similarity or Pearson correlation coefficient.

[0020] In the above solution, cosine similarity has the advantages of being simple to implement and fast in calculation speed. Pearson correlation coefficient has the advantages of being simple, intuitive, and easy to interpret.

[0021] Preferably, the edges of the scraper chain image are obtained by the Canny edge detection algorithm.

[0022] In the above solution, the Canny edge detection algorithm can effectively suppress image noise and reduce the influence of noise on edge detection.

[0023] Preferably, the colors of both the scraper chain and the sprocket are any one of red, blue, and green, and the scraper chain and the sprocket have different colors.

[0024] Preferably, each reference pixel point is an eight-neighborhood pixel point or a four-neighborhood pixel point.

[0025] Preferably, the scraper chain image is obtained by an image acquisition device provided at the feeding end of the scraper conveyor.

[0026] In the above solution, setting the image acquisition device in this way can prevent the material from blocking the scraper chain and causing the situation where the scraper chain image cannot be accurately obtained, and it does not affect the normal operation of the scraper conveyor.

[0027] Preferably, the value range of the preset duration is 8 to 12 seconds.

[0028] Preferably, the value range of the threshold is [0.5, 0.7].

[0029] The beneficial effects are as follows:

[0030] In the solution of the present invention, by introducing the tendency degree of pixel points, the scraper chain and the sprocket in the scraper chain image can be accurately distinguished, which is convenient for subsequent calculation of the abnormality degree of the scraper chain image. Moreover, by comparing the similarity of the horizontal features and the similarity of the vertical features in the scraper chain image, the abnormality degree of the scraper chain image can be comprehensively and accurately reflected, and thus the fault detection result is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 is a flowchart of the steps of the scraper chain fault detection method for a scraper conveyor according to an embodiment of the present invention;

[0032] Figure 2 is a flowchart of the steps of obtaining the abnormality degree of the scraper chain image according to an embodiment of the present invention;

[0033] Figure 3 is a distribution diagram of any pixel point and its reference pixel point in the scraper chain image according to an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The embodiments described below by referring to the drawings are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0035] AsFigure 1 As shown in the figure, the present invention provides a method for detecting faults of a scraper chain for a scraper conveyor, which includes the following steps:

[0036] S1. Collect images of the scraper chain at each moment in the rotating state.

[0037] The scraper conveyor includes a sprocket, a scraper chain, and a scraper. When the scraper conveyor works, the sprocket rotates and drives the scraper chain to move, and the scraper chain drives the scraper to move to realize the conveying of materials. When collecting the images of the scraper chain, the image acquisition device is fixedly arranged relative to the scraper conveyor, and through the movement of the scraper chain, the images of the scraper chain at each moment are collected. Therefore, in the present invention, the scraper conveyor can collect the images of the scraper chain without stopping, which does not affect the normal operation of the scraper conveyor and avoids the economic losses caused by stopping.

[0038] Since the top of the scraper chain will be covered by materials during the process of conveying materials, and thus the images of the scraper chain cannot be obtained, the image acquisition device can be arranged at the feeding end of the scraper conveyor. At this time, the scraper chain is in the state of unloading materials and not yet feeding materials. Due to the lack of material occlusion, good image acquisition effects can be achieved.

[0039] The image acquisition device can be a camera. By setting the photographing frequency of the camera, the images of the scraper chain at different moments can be collected. Of course, the image acquisition device can also be a video recorder, and the images of the scraper chain at different moments can be intercepted from the obtained video.

[0040] Since the scraper chain is driven by the sprocket, the images of the scraper chain collected include the scraper chain and the sprocket.

[0041] S2. Calculate the degree of abnormality of the images of the scraper chain at each moment within a preset time period. When the degrees of abnormality of the images of the scraper chain at all moments within the preset time period are greater than the threshold value, it is determined that the scraper chain has a fault.

[0042] When the scraper conveyor is working, the scraper chain is constantly moving. The scraper chain corresponding to the image acquisition device has different states at different times, that is, the angles of the scraper chain are different. Therefore, the scraper chain images collected at each time are also different. Due to the structural repeatability of the scraper chain, two scraper chain images separated by a certain time interval have a high similarity, that is to say, the change of the scraper chain image has periodicity. Under normal circumstances, the scraper chains at two times separated by a cycle have the same state, so the similarity of the corresponding two scraper chain images is high. When the scraper chain fails at a certain moment, for example, the scraper chain deviates left and right on the sprocket or the scraper chain breaks along its length direction, the similarity between the corresponding scraper chain image and the scraper chain image separated by a cycle will be low, that is, the abnormal degree of the corresponding scraper chain image is high. Therefore, it is possible to judge whether the scraper chain has a fault by calculating the abnormal degree of the scraper chain images at each time. Specifically, step S2 includes the following steps:

[0043] S21. Obtain the abnormal degree of the scraper chain images at each time.

[0044] The abnormal degree is inversely correlated with both the similarity of the horizontal features and the similarity of the vertical features between the scraper chain image at the current time and the scraper chain image at its set time. Among them, the horizontal feature represents the tendency degree of the pixel points in each row of the scraper chain image. By comparing the similarity of the horizontal features of the scraper chain images at two times, it can be reflected whether there are cracks or even fractures in the scraper chain in its length direction. The vertical feature represents the tendency degree of the pixel points in each column of the scraper chain image. By comparing the similarity of the vertical features of the scraper chain images at two times, it can be reflected whether the scraper chain is offset in the left-right direction. The horizontal feature is a row mean sequence composed of the means of the tendency degrees of the pixel points in each row of the scraper chain image, and the vertical feature is a column mean sequence composed of the means of the tendency degrees of the pixel points in each column of the scraper chain image. Therefore, as Figure 2 shown, step S21 also includes the following steps:

[0045] S211. Obtain the tendency degree of each pixel point.

[0046] The tendency degree represents the degree to which the corresponding pixel point tends to be located at the scraper chain in the scraper chain image. The tendency degree is positively correlated with the value obtained by normalizing the difference between the pixel values of the corresponding pixel point in the first color channel and the second color channel. The first color channel is the color channel with the largest pixel value of the pixel points in the color of the scraper chain, and the second color channel is the color channel with the largest pixel value of the pixel points in the color of the sprocket.

[0047] As is known to those skilled in the art, color images usually adopt the RGB color model. The RGB color model means that the color of any pixel point in a color image is formed by superimposing different pixel values of the red channel, green channel, and blue channel. For example, if the color of the scraper chain tends to be blue, then the color channel with the largest pixel value among the pixel points in the color of the scraper chain is the blue channel, that is, the first color channel is the blue channel. If the color of the sprocket tends to be red, then the color channel with the largest pixel value among the pixel points in the color of the sprocket is the red channel, that is, the second color channel is the red channel. In the present invention, the colors of the scraper chain and the sprocket are each any one of red, blue, and green, and the scraper chain and the sprocket have different colors.

[0048] Since the first color channel is the color channel with the largest pixel value among the pixel points in the color of the scraper chain, the pixel value of the scraper chain in the first color channel of the scraper chain image is greater than that in the second color channel. Moreover, the greater the difference between the pixel value of the pixel point in the first color channel and that in the second color channel, the more likely it is that the pixel point is located at the scraper chain in the scraper chain image, that is, the greater the inclination degree of the pixel point, and the smaller the difference, the less likely it is that the pixel point is located at the scraper chain in the scraper chain image, that is, the smaller the inclination degree of the pixel point.

[0049] The inclination degree is also positively correlated with the value obtained by normalizing the difference between the pixel values of each reference pixel point in the first color channel and the second color channel. The reference pixel point is the first pixel point after the multiple rays radiating outward from the corresponding pixel point pass through the nearest edge.

[0050] In one embodiment, as Figure 3 shown, each reference pixel point is an eight-neighborhood pixel point, that is, the eight-neighborhood pixel points are respectively located on eight rays radiating outward from the corresponding pixel point. And the angle between the ray where the first reference pixel point is located and the horizontal right vector is 0 degrees, the angle between the ray where the second reference pixel point is located and the horizontal right vector is 45 degrees, the angle between the ray where the third reference pixel point is located and the horizontal right vector is 90 degrees, the angle between the ray where the fourth reference pixel point is located and the horizontal right vector is 135 degrees, the angle between the ray where the fifth reference pixel point is located and the horizontal right vector is 180 degrees, the angle between the ray where the sixth reference pixel point is located and the horizontal right vector is 225 degrees, the angle between the ray where the seventh reference pixel point is located and the horizontal right vector is 270 degrees, and the angle between the ray where the eighth reference pixel point is located and the horizontal right vector is 315 degrees. Of course, each reference pixel point can also be a four-neighborhood pixel point.

[0051] The edges of the scraper chain image are obtained by the Canny edge detection algorithm. The Canny edge detection algorithm can effectively suppress image noise and reduce the influence of noise on edge detection. Of course, the Laplace algorithm can also be used for edge detection.

[0052] The reason why the present invention also uses reference pixel points to characterize the tendency degree of pixel points is as follows: when the scraper conveyor is actually working, protective paint of different colors is usually applied on the scraper chain and the sprocket. The color of the protective paint is usually any one of red, blue, and green. As the scraper conveyor is used, the protective paint on the scraper chain will wear out, and even be partially missing, exposing the metal color of the scraper chain. This part of the scraper chain is called the missing part. In order to avoid a relatively low tendency degree of the pixel points in the missing part being calculated, it is necessary to correct the tendency degree of the pixel points in the missing part.

[0053] The reference pixel points are the part adjacent to the outside of the missing part. The greater the difference in the pixel values of the reference pixel points in the first color channel and the second color channel, the greater the possibility that the reference pixel points are located at the scraper chain, and then the greater the possibility that the corresponding pixel points are located in the missing part. It is necessary to correct the tendency degree of this pixel point to increase the tendency degree of this pixel point.

[0054] Therefore, the tendency degree of the th pixel point in the scraper chain image at the th moment is:

[0055] ;

[0056] In the formula, is the pixel value of the th pixel point in the first color channel in the scraper chain image at the th moment, is the pixel value of the th pixel point in the second color channel in the scraper chain image at the th moment, is the pixel value of the th reference pixel point of the th pixel point in the first color channel in the scraper chain image at the th moment, is the pixel value of the th reference pixel point of the th pixel point in the second color channel in the scraper chain image at the th moment, is the normalization function, is the adjustment parameter, , is the th moment of the scraper chain image, and the The total number of reference pixels for one pixel point, is an exponential function with the natural constant as the base.

[0057] In the above formula, the differences between the pixel values of the pixel point and the reference pixel point in the first color channel and the second color channel are both divided by 255. This is to prevent the normalized value from approaching 0 or -1 when the difference is too large or too small, so that it cannot well reflect and correct the tendency degree. Set the adjustment parameter to make the correction strength greater. Preferably, is 2.

[0058] is a normalization function, which can map the calculation result to the range of 0 to 1. Of course, the min-max normalization function can also be used for normalization processing.

[0059] In the present invention, the tendency degree of the pixel point is not only characterized by the difference between the pixel values of the pixel point in the first color channel and the second color channel, but also corrected by the differences between the pixel values of each reference pixel point in the first color channel and the second color channel, preventing the calculated tendency degree from being too small due to the wear of the protective paint on the scraper chain, resulting in inaccurate calculation results.

[0060] S212. According to the tendency degree of each pixel point, obtain the horizontal feature and vertical feature of the scraper chain image at each moment.

[0061] Take the row mean sequence formed by the means of the tendency degrees of the pixel points in each row of the scraper chain image as the horizontal feature of the corresponding scraper chain image, and take the column mean sequence formed by the means of the tendency degrees of the pixel points in each column of the scraper chain image as the vertical feature of the corresponding scraper chain image.

[0062] S213. According to the similarity of the horizontal feature and the similarity of the vertical feature of the scraper chain image, obtain the abnormality degree of the scraper chain image at each moment.

[0063] Since the abnormality degree is inversely correlated with both the similarity of the horizontal feature and the similarity of the vertical feature between the scraper chain image at the current moment and the scraper chain image at the previously set moment. Therefore, it is necessary to obtain the set moment of the scraper chain image, and the set moment is the moment closest to the maximum value of the similarity of the horizontal feature of the scraper chain image at the current moment. The time interval between the set moment and the current moment can also be called the period of change of the scraper chain image.

[0064] In one embodiment, recalculate the set moment of the scraper chain image every 2 to 5 minutes to adapt to the situation where the conveying speed of the scraper conveyor changes.

[0065] In this step, when the similarity of the horizontal features of the scraper chain images at the current moment and the set moment reaches the maximum value, the scraper chains at the two moments are in the same state, so they can be compared. Moreover, the set moment can be automatically calculated, reducing the dependence on manual adjustment, improving the adaptability of the fault detection method, and enabling the scraper chain fault detection to adapt to different conveying speeds, scenario changes, and equipment changes.

[0066] Therefore, the degree of abnormality of the scraper chain image at the th moment is :

[0067] ;

[0068] In the formula, is the similarity of the horizontal features of the scraper chain images at the th moment and the th moment, is the similarity of the vertical features of the scraper chain images at the th moment and the th moment. The th moment is the set moment, and the th moment is the moment closest to the maximum value of the similarity of the horizontal features of the scraper chain image at the th moment. is the exponential function with the natural constant as the base.

[0069] In this formula, and are each added by 1 and then divided by 2 for the purpose of normalization, facilitating subsequent calculations.

[0070] In one embodiment, the similarity of the horizontal features and the similarity of the vertical features of the scraper chain image are both calculated using the cosine similarity. The cosine similarity has the advantages of being simple to implement and fast in calculation speed.

[0071] In another embodiment, the similarity can also be represented by the Pearson correlation coefficient. The Pearson correlation coefficient has the advantages of being simple, intuitive, and easy to interpret.

[0072] In step S21, the degree of abnormality of the scraper chain image is characterized by the similarity of the horizontal features of the scraper chain image, which can reflect whether there are cracks or fractures along the length of the scraper chain. Characterized by the similarity of the vertical features of the scraper chain image, it can reflect whether the scraper chain is shifted left or right. Therefore, it can comprehensively and accurately reflect the degree of abnormality of the scraper chain image.

[0073] S22. Compare the abnormality degrees of the scraper chain images at each moment within the preset duration with the threshold value respectively. When the abnormality degrees of all scraper chain images within the preset duration are greater than the threshold value, it is determined that the scraper chain has a fault.

[0074] In step S22, by setting the preset duration and calculating the abnormality degree of the scraper chain images within a period of time, accidental errors can be reduced, making the calculation result more accurate. The value range of the preset duration is 8 to 12 seconds. Preferably, the preset duration is 10 seconds. Of course, the size of the preset duration can be adjusted according to needs.

[0075] The value range of the threshold is [0.5, 0.7]. Preferably, the threshold is 0.6. Of course, the size of the threshold can be adjusted according to needs.

[0076] After it is determined that the scraper chain has a fault, the scraper conveyor can be controlled to stop running, and then the staff can check or repair the scraper chain to prevent safety accidents.

[0077] In the method for detecting faults of the scraper chain of the scraper conveyor of the present invention, by calculating the differences in the pixel values of each pixel point in the first color channel and the second color channel of the scraper chain image, as well as the differences in the pixel values of the corresponding reference pixel points in the first color channel and the second color channel, the inclination degree of each pixel point is obtained, and the scraper chain and the sprocket in the scraper chain image can be accurately separated, thus facilitating the detection of faults of the scraper chain. And by comparing the similarities of the horizontal features and the vertical features in the scraper chain image, the abnormality degree of the scraper chain image at each moment is obtained, and then according to the abnormality degrees of all scraper chain images within the preset duration, it is determined whether the scraper chain has a fault, which can not only comprehensively and accurately reflect the abnormality degree of the scraper chain image, but also avoid misjudgment caused by accidental errors, making the fault detection result more reliable.

[0078] Although this specification has shown and described multiple embodiments of the present invention, it is obvious to those skilled in the art that such embodiments are provided only by way of example. Those skilled in the art will think of many changes, alterations and alternative ways without departing from the spirit and idea of the present invention. It should be understood that various alternative solutions to the embodiments of the present invention described herein can be adopted in the process of practicing the present invention.

Claims

1. A method for detecting faults of a scraper chain for a scraper conveyor, characterized in that, The method includes the following steps: collecting the images of the scraper chain at each moment in the rotating state; calculating the degree of abnormality of the images of the scraper chain at each moment within a preset time period, and determining that the scraper chain has a fault when the degrees of abnormality of the images of the scraper chain at all moments within the preset time period are greater than a threshold value; wherein, the degree of abnormality is inversely correlated with both the similarity of the horizontal features and the similarity of the vertical features between the image of the scraper chain at the current moment and the image of the scraper chain at a previously set moment; the horizontal feature is a row mean sequence formed by the means of the inclination degrees of the pixel points in each row of the scraper chain image, and the vertical feature is a column mean sequence formed by the means of the inclination degrees of the pixel points in each column of the scraper chain image; the inclination degree is positively correlated with the value obtained by normalizing the difference between the pixel values of the corresponding pixel point in the first color channel and the second color channel, and is positively correlated with the value obtained by normalizing the difference between the pixel values of each reference pixel point in the first color channel and the second color channel; the reference pixel point is the first pixel point after a plurality of rays radiating outward from the corresponding pixel point pass through the nearest edge; the first color channel is the color channel with the largest pixel value among the colors of the pixel points of the scraper chain, and the second color channel is the color channel with the largest pixel value among the colors of the pixel points of the sprocket; The abnormality level of the scraper chain image at the following time is: ； Wherein, is the similarity of the lateral features of the scraper chain images at the -th moment and the -th moment, is the similarity of the longitudinal features of the scraper chain images at the -th moment and the -th moment. The -th moment is the set moment, and the -th moment is the moment closest to the maximum value of the similarity of the lateral features of the scraper chain image at the -th moment, is an exponential function with the natural constant as the base.

2. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that, The tendency degree of the th pixel point in the scraper chain image at the th moment is as follows: ； In the formula, is the pixel value of the th pixel point in the th scraper chain image at the th moment in the first color channel, is the pixel value of the th pixel point in the th scraper chain image at the th moment in the second color channel, is the pixel value of the th reference pixel point of the th pixel point in the th scraper chain image at the th moment in the first color channel, is the pixel value of the th reference pixel point of the th pixel point in the , is the th total number of reference pixel points of the th pixel point in the th scraper chain image at the th moment, and is the exponential function with the natural constant as the base.

3. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that, The similarity is calculated by using cosine similarity or Pearson correlation coefficient.

4. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that, The edge of the scraper chain image is obtained by using the Canny edge detection algorithm.

5. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that The colors of both the scraper chain and the sprocket are any one of red, blue, and green, and the scraper chain and the sprocket are of different colors.

6. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that, Each reference pixel point is an eight-neighborhood pixel point or a four-neighborhood pixel point.

7. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that, The scraper chain image is collected by an image acquisition device arranged at the feeding end of the scraper conveyor.

8. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that, The value range of the preset time period is 8 to 12 seconds.

9. The scraper chain fault detection method for a scraper conveyor according to claim 1, characterized in that, The value range of the threshold value is [0.5, 0.7].

Images