CN115684174A - Agricultural product transportation conveyor belt safe operation monitoring method - Google Patents
Agricultural product transportation conveyor belt safe operation monitoring method Download PDFInfo
- Publication number
- CN115684174A CN115684174A CN202211281903.2A CN202211281903A CN115684174A CN 115684174 A CN115684174 A CN 115684174A CN 202211281903 A CN202211281903 A CN 202211281903A CN 115684174 A CN115684174 A CN 115684174A
- Authority
- CN
- China
- Prior art keywords
- conveyor belt
- suspected
- monitoring
- gradient
- data information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 71
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000011156 evaluation Methods 0.000 claims abstract description 66
- 238000009826 distribution Methods 0.000 claims abstract description 32
- 238000000605 extraction Methods 0.000 claims abstract description 26
- 230000035772 mutation Effects 0.000 claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 24
- 230000004044 response Effects 0.000 claims abstract description 18
- 238000001914 filtration Methods 0.000 claims abstract description 17
- 239000011159 matrix material Substances 0.000 claims abstract description 14
- 230000008569 process Effects 0.000 claims description 27
- 230000004927 fusion Effects 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 15
- 230000001186 cumulative effect Effects 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 4
- 238000003708 edge detection Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000009825 accumulation Methods 0.000 claims 1
- 238000001514 detection method Methods 0.000 description 17
- 230000011218 segmentation Effects 0.000 description 14
- 230000000694 effects Effects 0.000 description 8
- 230000005856 abnormality Effects 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000001154 acute effect Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 239000003245 coal Substances 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000005272 metallurgy Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 238000006748 scratching Methods 0.000 description 1
- 230000002393 scratching effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Control Of Conveyors (AREA)
Abstract
The invention relates to the technical field of conveyor belt monitoring, in particular to a method for monitoring the safe operation of an agricultural product transportation conveyor belt; acquiring a gray image corresponding to the conveyor belt; filtering the gray level image in multiple directions by using filtering kernels with different scales to obtain multiple feature extraction images, and fusing the feature extraction images to obtain a saliency map; extracting suspected crack areas in the saliency map, calculating gradient distribution characteristics, random indexes and intensity response values corresponding to the suspected crack areas to further obtain evaluation indexes, and obtaining real cracks based on the evaluation indexes; constructing a monitoring matrix according to the data information, acquiring mutation points of each row of data information, and calculating an abnormal evaluation value when the number of the mutation points of any row of data information is greater than a set number; and calculating an operation state index according to the evaluation index and the abnormal evaluation value corresponding to each real crack, and monitoring the conveying belt based on the operation state index. The invention can accurately realize the purpose of monitoring the conveyor belt.
Description
Technical Field
The invention relates to the technical field of conveyor belt monitoring, in particular to a method for monitoring the safe operation of an agricultural product transportation conveyor belt.
Background
The conveying belt for transportation is widely applied to multiple industries such as coal mines, chemical industry, agriculture and metallurgy, has the main advantages of large conveying capacity, large adaptability of transportation distance, economy, safety, high reliability and the like, and is one of main transportation devices for agricultural product transportation in the agricultural production and transportation process. In the working process of the conveyor belt, accidents such as longitudinal tearing of the conveyor belt, large-amplitude vibration of the conveyor belt and the like are caused by unexpected conditions such as blocking of metal object clamps, clamping and pressing of a roller and a carrier roller, scratching of waste rocks and the like, and the stability of the conveyor belt in operation is influenced. The belt of conveyer belt can produce the condition emergence that wearing and tearing are ageing to lead to the fracture, and the conveyer belt is once broken and can seriously influence the agricultural product transportation in time repair and change, takes place unpredictable danger. Therefore, it is important how to obtain the crack of the conveyor belt and to accurately monitor the running process of the conveyor belt.
The method for acquiring the cracks of the conveyor belt in the prior art comprises the following steps: the surface image of the conveyor belt is obtained, the crack of the conveyor belt is obtained by utilizing an image processing technology, namely, the crack is obtained by carrying out threshold segmentation on the image, and therefore the monitoring of the running process of the conveyor belt is realized. The method has high requirements on the threshold value, and once the accuracy of threshold value selection is not enough, the crack of the conveyor belt cannot be found in time, and the accurate monitoring of the conveyor belt cannot be realized.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a method for monitoring the safe operation of an agricultural product transportation conveyor belt, which adopts the following technical scheme:
acquiring a surface image of a conveyor belt, and preprocessing the surface image of the conveyor belt to obtain a gray image corresponding to the conveyor belt; filtering the gray level image in multiple directions by using filtering kernels with different scales respectively to obtain multiple feature extraction images; fusing a plurality of feature extraction graphs to obtain a saliency map;
extracting a straight line segment area in the saliency map, marking the straight line segment area as a suspected crack area, and obtaining a gradient distribution characteristic corresponding to each suspected crack area according to a gradient amplitude and a gradient direction angle corresponding to each suspected crack pixel point in the suspected crack area;
obtaining random indexes corresponding to the suspected crack areas according to the width and the length of the minimum external rectangle corresponding to each suspected crack area and the distance from all the suspected crack pixel points to the edge line of the conveyor belt;
recording the gray average value of all suspected crack pixel points corresponding to each suspected crack area as an intensity response value, obtaining an evaluation index corresponding to each suspected crack area based on the gradient distribution characteristics, the random index and the intensity response value corresponding to each suspected crack area, and recording the suspected crack area corresponding to the evaluation index larger than a set threshold value as a real crack;
acquiring data information corresponding to different monitoring parameters of a conveyor belt in the running process, constructing a monitoring matrix based on the data information, and acquiring mutation points corresponding to each row of data information;
and calculating the running state index of the conveyor belt according to the evaluation index corresponding to each real crack and the abnormal evaluation value, and monitoring the conveyor belt based on the running state index.
Preferably, the method for obtaining the saliency map by fusing the plurality of feature extraction maps specifically comprises the following steps: firstly, carrying out position alignment operation on feature extraction graphs corresponding to different directions under the same scale, and calculating a pixel mean value corresponding to the position according to pixel values of pixel points corresponding to the same position to obtain a fusion feature extraction graph corresponding to the scale; and then respectively carrying out significance processing on the fusion feature images corresponding to the non-scale scales to obtain fusion significance images corresponding to different scales, further distributing weights to each fusion significance image, and carrying out weighting processing on all fusion significance images according to the weights to obtain the significance images.
Preferably, the method for acquiring the gradient distribution characteristics comprises the following steps: recording the gradient amplitude and the gradient direction angle corresponding to each suspected crack pixel point as gradient information, forming a gradient information set corresponding to each suspected crack region by the gradient information corresponding to all the suspected crack pixel points corresponding to each suspected crack region, recording the gradient information with the same value as the same gradient information, calculating the ratio of the frequency of each gradient information appearing in the gradient information set to the quantity of all the gradient information in the gradient information set, and determining the gradient distribution characteristics corresponding to each suspected crack region according to the variance value of the ratio and all the gradient direction angles in the gradient information set.
Preferably, the method for acquiring the random index comprises:
and calculating the cumulative sum of the distances from all the suspected crack pixel points in each suspected crack area to the edge line of the conveyor belt, calculating the ratio of the width to the length of the minimum circumscribed rectangle, and determining a random index according to the cumulative sum and the ratio.
Preferably, the evaluation index is:
wherein, P k Ran is an evaluation index corresponding to the suspected crack region k k Is a random index corresponding to the suspected crack region k,
the intensity response value corresponding to the suspected crack area k is obtained; gray k ' is a value after normalization of gradient distribution characteristics corresponding to a suspected fracture region k, theta k Forming an included angle between a straight line corresponding to the suspected crack area k and the edge line of the conveyor belt; e is a natural constant.
Preferably, the method for obtaining the mutation point corresponding to each row of data information includes: and acquiring edge values corresponding to all data information in the monitoring matrix through a horizontal edge detection operator, recording the data information with the edge value of 1 as a mutation point, and further acquiring the mutation point corresponding to each row of data information.
Preferably, the abnormality evaluation value is:
wherein z is c For the abnormal evaluation value corresponding to the c-th row data information, h c,max Is the maximum value corresponding to the c-th row of data information, h c,min Is the minimum value, sigma, corresponding to the c-th row of data information c The variance corresponding to the information of the c-th row of data; exp (. Cndot.) is an exponential function based on a natural constant e.
Preferably, the operating state index is a sum of all abnormal evaluation values and an accumulated sum of evaluation indexes corresponding to all real fractures.
The embodiment of the invention at least has the following beneficial effects:
according to the method, filtering processing is carried out on the gray level image in multiple directions by using filtering kernels with different scales to obtain multiple feature extraction images, and then the feature extraction images are fused to obtain a saliency map; and extracting suspected crack areas in the saliency map, calculating gradient distribution characteristics, random indexes and intensity response values corresponding to the suspected crack areas to further obtain evaluation indexes, and acquiring real cracks based on the evaluation indexes. The salient image can clearly reflect texture information in the gray image, and then the texture on the surface of the conveyor belt can be accurately obtained. The detection precision of the real crack is improved in the subsequent operation process, and the crack area is convenient to identify. The calculation of the evaluation index combines the gradient distribution characteristics, the random index and the strength response value, considers factors in various aspects, and can accurately acquire the real crack. Meanwhile, the invention also calculates the abnormal evaluation value corresponding to each row of data information by constructing a monitoring matrix, calculates the running state index according to the evaluation index and the abnormal evaluation value corresponding to each real crack, and monitors the conveyor belt based on the running state index. The conveyor belt monitoring method and the conveyor belt monitoring system have the advantages that the conveyor belt is monitored not only from the aspect of cracks on the surface of the conveyor belt, but also from the aspect of data information corresponding to monitoring parameters of the conveyor belt, the purpose of automatically monitoring the conveyor belt can be accurately achieved, and the monitoring precision is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flow chart of steps of an embodiment of a method for monitoring the safe operation of an agricultural product transport conveyor belt according to the present invention.
Detailed Description
To further explain the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the proposed solution, its specific implementation, structure, features and effects will be made with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
Referring to fig. 1, a flow chart of steps of a method for monitoring the safe operation of a agricultural product transport conveyor according to an embodiment of the present invention is shown, the method comprising the steps of:
step 1, acquiring a surface image of a conveyor belt, and preprocessing the surface image of the conveyor belt to obtain a gray image corresponding to the conveyor belt; filtering the gray level image in multiple directions by using filtering kernels with different scales respectively to obtain multiple feature extraction images; and fusing the plurality of feature extraction images to obtain a saliency map.
Specifically, firstly, the image of the conveyor belt is acquired by a camera, and a surface image of the conveyor belt is acquired and used for detecting the surface condition of the conveyor belt. The camera needs to be ensured to be capable of completely collecting the image of the surface of the conveyor belt in the process of collecting the image of the conveyor belt, so that the camera is arranged at the initial end of the conveyor belt, the image of the surface of the conveyor belt is collected in real time in the process of running the conveyor belt, the position arrangement of the camera, the setting of the parameters of the camera and the like are carried out, and an implementer sets the camera according to the actual situation; meanwhile, the time interval between two adjacent detection moments and the time interval between two adjacent time periods of the camera for acquiring the surface images of the conveyor belt are consistent with the time interval between two adjacent detection moments and the time interval between two adjacent time periods corresponding to the data information corresponding to different monitoring parameters of the follow-up acquisition conveyor belt in the operation process. The splicing and fusing method is a known technology, and an implementer can select a corresponding method according to specific situations.
And then analyzing the obtained spliced image to analyze the crack problem on the surface of the conveyor belt, identifying and extracting crack regions so as to represent the abnormal condition of the surface of the conveyor belt, regarding the acquired conveyor belt image, considering that other irrelevant regions interfere with the identification of the crack regions, and considering that the distribution condition of the subsequent analysis crack regions is in certain correlation with the position information of two edges of the conveyor belt, firstly extracting a target region of the spliced image, namely obtaining a gray image of the conveyor belt, specifically, carrying out gray processing on the spliced image to obtain a gray image corresponding to the spliced image, identifying the conveyor belt region through a trained semantic segmentation network, and further obtaining the gray image corresponding to the conveyor belt. Artificially setting the pixel point of the conveyor belt region to be 1, setting the gray value of other pixel points to be 0 to obtain a label image, training the semantic segmentation network by utilizing a large number of label images and gray images corresponding to the spliced image to obtain the trained semantic segmentation network, wherein the specific semantic segmentation process of the semantic segmentation network is the prior known technology, is not in the protection range of the invention and is not repeated.
Further, in order to avoid the problems of fuzzy and discontinuous edge extraction in the semantic segmentation process, in this embodiment, based on a semantic segmentation effect map obtained through a semantic segmentation network (the semantic segmentation effect map is a binary image, and a pixel point with a pixel value of 1 in the semantic segmentation effect map is a pixel point corresponding to a conveyor belt), a set of pixel points corresponding to the upper edge and the lower edge of the conveyor belt is obtained and recorded as U u 、U d And respectively fitting two straight lines through the position information of the pixel points in the respective pixel point set to obtain an upper edge line Y of the conveyor belt u And lower edge line Y d In this embodiment, fitting is performed by using RANSAC algorithm, and then the upper edge line Y in the semantic segmentation effect map is obtained u And a lower edge line Y d And setting the pixel value of the included pixel point to be 1, finishing the correction of the semantic segmentation effect graph, and multiplying the corrected semantic segmentation effect graph serving as a mask with the gray level image to obtain the gray level image corresponding to the conveyor belt. The RANSAC algorithm is well known and will not be described in detail.
In order to extract the crack region on the surface of the conveyor belt more quickly and accurately, the embodiment respectively uses filtering kernels with different scales to filter the gray level image in multiple directions to obtain multiple feature extraction images; and fusing the multiple feature extraction images to obtain a saliency map.
Specifically, when a Gabor filter is used to filter a grayscale image, a feature extraction map can be obtained in each direction for a filter kernel of one scale, for example, a feature extraction map can be obtained in a direction of 0 ° for a filter kernel of 3 × 3, and similarly, a feature extraction map can also be obtained in a direction of 45 ° for a filter kernel of 3 × 3. In this embodiment, 3 filtering kernels with different scales are used to perform filtering processing on the gray-scale image in 8 directions, where the scales of the filtering kernels are 3 × 3, 5 × 5, and 7 × 7, and 8 directions correspond to 8 neighborhood directions of the pixel points, that is, 24 feature extraction graphs are obtained altogether. The implementer can select the size of the filter kernels, the number of the filter kernels and the number of the directions according to actual conditions in the actual operation process. The filtering process of the gray image by using the Gabor filter is a known technique, and is not within the protection scope of the present invention, and is not described in detail.
The method for obtaining the saliency map by fusing the multiple feature extraction maps specifically comprises the following steps: firstly, carrying out position alignment operation on feature extraction graphs corresponding to different directions under the same scale, and calculating a pixel mean value corresponding to the position according to pixel values of pixel points corresponding to the same position to obtain a fusion feature extraction graph corresponding to the scale; and then respectively carrying out significance processing on the fusion feature maps corresponding to the non-scale to obtain fusion significance maps corresponding to different scales, further distributing weights to each fusion significance map, and carrying out weighting processing on all fusion significance maps according to the weights to obtain the significance maps.
Specifically, an Itti significance analysis algorithm is selected to perform significance processing on the fusion feature map, and an implementer may select other significance analysis algorithms, such as an AC algorithm, an HC algorithm, and the like. The significance analysis algorithm is an existing known algorithm and is not elaborated in detail; then, when the weights are distributed to all the fusion saliency maps, the same weights are distributed, namely three fusion saliency maps are distributed, the weight distributed to each fusion saliency map is 1/3, and in the actual operation process, different weights can be set according to different information contained in each fusion saliency map.
It should be noted that, by performing filtering processing on the grayscale image in multiple directions by using filtering kernels of different scales, texture information in the grayscale image can be extracted, and a texture of the surface of the conveyor belt can be obtained. The obtained texture information can be enhanced through remarkable processing, the detection precision of the crack region is improved in the subsequent operation process, and the crack region is convenient to identify.
And 2, extracting a straight line segment area in the saliency map, marking the straight line segment area as a suspected crack area, and obtaining a gradient distribution characteristic corresponding to each suspected crack area according to a gradient amplitude and a gradient direction angle corresponding to each suspected crack pixel point in the suspected crack area.
Because most cracks are in a straight line state, a Hough line detection algorithm is adopted to extract a straight line segment area in the saliency map and mark the straight line segment area as a suspected crack area, the Hough line detection algorithm is a known technology, and the specific process is not repeated.
Considering that in an actual analysis scene, the surface of the conveyor belt still has more grains formed due to uneven distribution of the surface of the conveyor belt and partial scratches generated in the operation process, the scratches do not actually bring practical damage to the conveyor belt, the scratches and the grains formed due to uneven distribution of the surface of the conveyor belt are collectively referred to as false cracks, and because the similarity between the false cracks and the real cracks is high, only a Hough line detection algorithm cannot judge which are the real cracks and which are the false cracks, so that the suspected crack region is further analyzed.
Specifically, the gradient distribution characteristics corresponding to each suspected crack area are obtained according to the gradient amplitude and the gradient direction angle corresponding to each suspected crack pixel point in the suspected crack area.
The method for acquiring the gradient distribution characteristics comprises the following steps: recording the gradient amplitude and the gradient direction angle corresponding to each suspected crack pixel point as gradient information
i＝1,2,…,N k Wherein
the gradient amplitude corresponding to the ith suspected crack pixel point in the suspected crack area k,
is the gradient direction angle, N, corresponding to the ith suspected crack pixel point in the suspected crack area k k The total number of suspected crack pixel points in the suspected crack area k is obtained. The method for acquiring the gradient amplitude and the gradient direction angle corresponding to the suspected crack pixel point is a known technology; gradient information corresponding to all the suspected crack pixel points corresponding to each suspected crack area forms a gradient information set corresponding to the suspected crack area, the gradient information with the same value is recorded as the same gradient information, and each gradient information is calculatedDetermining the gradient distribution characteristics corresponding to each suspected crack area according to the ratio of the frequency of gradient information appearing in a gradient information set to the quantity of all gradient information in the gradient information set and the variance value of all gradient direction angles in the gradient information set; the formula for the gradient profile is expressed as:
wherein, gray k Is the gradient distribution characteristic corresponding to the suspected fracture region k,
is the variance value, w, of all gradient direction angles in the suspected crack region k v Is the ratio of the frequency of the V-th gradient information in the gradient information set to the quantity of all gradient information in the gradient information set, V k The number of types of gradient information; e is a natural constant; lg w v Is based on a natural constant of 10 v The logarithm of (d).
Furthermore, the gradient distribution characteristics are subjected to normalization processing, the values of the gradient distribution characteristics are ensured to be in an interval (0, 1), and the evaluation indexes of the suspected crack areas are conveniently calculated in a follow-up mode.
Considering that when a real crack appears on the surface of the conveyor belt, the gradient information of the crack pixel points corresponding to the crack in the same range is kept consistent, that is, the number of the types of the gradient information in the corresponding gradient information set is not large, and the ratio of the frequency of each type of gradient information appearing in the gradient information set to the number of all the gradient information in the gradient information set should be large, and at the same time,
the fluctuation conditions of all suspected crack pixel points in the suspected crack area k can be represented, and the larger the fluctuation condition is, the more unlikely the crack is to be a real crack; based on the method, the gradient distribution characteristics corresponding to each suspected crack area are calculated and used as a parameter for subsequently judging whether each suspected crack area is a real crack or notConsidering the factors, when the gradient distribution characteristics are larger, the more complicated the gradient distribution condition in the suspected fracture area is, that is, the more inconsistent the gradient information is, the lower the possibility that the suspected fracture area is a real fracture is.
And 3, obtaining random indexes corresponding to the suspected crack areas according to the width and the length of the minimum external rectangle corresponding to the suspected crack areas and the distances from all the suspected crack pixel points to the edge lines of the conveyor belt.
The cracks grow randomly, scratches are defects formed at one time caused by acute, and cracks on the surface of the conveyor belt grow randomly, so that the random index is constructed for analyzing the defect distribution condition of a suspected crack area.
The method for acquiring the random index comprises the following steps: calculating the cumulative sum of the distances from all suspected crack pixel points in each suspected crack area to the edge line of the conveyor belt, calculating the ratio of the width to the length of the minimum circumscribed rectangle, and determining a random index according to the cumulative sum and the ratio, wherein the formula of the random index is expressed as follows:
wherein Ran is k Is a random index corresponding to the suspected crack area k, w k The width of the minimum circumscribed rectangle corresponding to the suspected crack area k; h is a total of k The length of the minimum circumscribed rectangle corresponding to the suspected crack area k; d is a radical of i,k The distance N from the ith suspected crack pixel point in the suspected crack area k to the edge line of the conveyor belt k The total number of suspected crack pixel points in the suspected crack area k is shown.
And 4, recording the gray average value of all the suspected crack pixel points corresponding to each suspected crack area as an intensity response value, obtaining an evaluation index corresponding to each suspected crack area based on the gradient distribution characteristics, the random index and the intensity response value corresponding to each suspected crack area, and recording the suspected crack area corresponding to the evaluation index larger than a set threshold value as a real crack.
When cracks appear on the surface of the conveyor belt, the cracks have a certain depth, and the gray value of the cracks is lower than that of defects such as scratches during image acquisition, so that the gray average value of all the suspected crack pixel points corresponding to each suspected crack area is calculated and recorded as an intensity response value, and the smaller the intensity response value, the higher the possibility that the corresponding suspected crack area is a real crack is.
And then obtaining the evaluation index corresponding to each suspected crack area based on the gradient distribution characteristics, the random index and the strength response value corresponding to each suspected crack area.
The evaluation indexes are as follows:
wherein, P k Ran is an evaluation index corresponding to a suspected crack region k k Is a random index corresponding to the suspected crack region k,
is suspected to be crackedThe intensity response value corresponding to the seam region k; gray k ' is a value normalized by the gradient distribution characteristic corresponding to the suspected crack region k, theta k Forming an included angle between a straight line corresponding to the suspected crack area k and an edge line of the conveyor belt; e is a natural constant.
The straight line corresponding to the suspected crack area k is obtained by fitting the suspected crack pixel points corresponding to the suspected crack area, and the fitting method is a known technology and is not repeated. Considering that the conveyor belt has certain tension in the running direction during running, the direction of most tearing cracks on the surface of the conveyor belt is generally vertical to the running direction of the conveyor belt, namely, the included angle theta formed by the straight line corresponding to the area introducing the suspected cracks and the edge line of the conveyor belt k And the evaluation index is calculated, so that the real crack can be more accurately obtained.
When the evaluation index is calculated, not only are the gradient distribution characteristics, the random index and the intensity response value corresponding to the suspected defect area combined, but also included angles formed by straight lines corresponding to the suspected crack area and edge lines of the conveyor belt are introduced, so that various factors are considered, and the real crack can be accurately obtained. Evaluation index Ran k Characterizing the probability that the suspected fracture area is a real fracture; according to the step 3, the larger the random index is, the higher the possibility that the corresponding suspected crack area is a real crack is, namely, the random index and the evaluation index show positive correlation; gray k The larger the value is, the more complicated the gradient distribution condition in the suspected fracture area is characterized, and the more inconsistent the gradient information is, the smaller the possibility that the corresponding suspected fracture area is a real fracture is, namely Gray k ' the evaluation index presents a negative correlation; intensity response value
The larger the intensity value is, the higher the gray value representing the suspected crack area is, which indicates that the corresponding suspected crack area is lower in possibility of being a real crack, namely, the intensity response value and the evaluation index present a negative correlation relationship.
And obtaining an evaluation index corresponding to each suspected crack region, comparing the evaluation index with a set threshold value, marking the suspected crack region corresponding to the evaluation index larger than the set threshold value as a real crack, marking the suspected crack region corresponding to the evaluation index smaller than or equal to the set threshold value as a false crack, and not participating in monitoring the conveying belt state in the subsequent process of the false crack.
In order to obtain a real crack more quickly in this embodiment, the evaluation index is normalized first, so that the value of the normalized evaluation index is 0 to 1, then the normalized evaluation index is compared with a set threshold, a suspected crack area corresponding to the normalized evaluation index being greater than the set threshold is recorded as a real crack, a suspected crack area corresponding to the normalized evaluation index being smaller than the set threshold is recorded as a false crack, wherein the value of the set threshold is 0.6, and in a specific implementation process, an implementer can adjust the value of the set threshold according to actual conditions.
And 5, acquiring data information corresponding to different monitoring parameters of the conveyor belt in the running process, constructing a monitoring matrix based on the data information, and acquiring mutation points corresponding to each row of data information, wherein when the number of the mutation points corresponding to any row is larger than a set number, the abnormal evaluation value corresponding to the row of data information is calculated according to the maximum value, the minimum value and the variance corresponding to the row of data information.
In order to analyze and recognize abnormal conditions of factors which are not easy to see in the running process of the conveyor belt, data information corresponding to different monitoring parameters of the conveyor belt in the running process is collected in real time, the monitoring parameters include but are not limited to conveyor belt frequency, conveyor belt tension, conveyor belt running power and the like, the data information corresponding to the different monitoring parameters is obtained by corresponding sensors or data collection equipment, the number of the monitoring parameters is recorded as C, and an implementer selects the value of C according to specific conditions. Considering that most of the collected data information is continuous and inconvenient to analyze, therefore, the collected data information is discretized, for each monitoring parameter, in this embodiment, the time interval corresponding to two adjacent detection times is set to be T =1s, the time interval corresponding to two adjacent time periods is set to be F =10min, that is, monitoring and identification of the running condition of the conveyor belt is performed every ten minutes, and s data information corresponding to each monitoring parameter is collected every 10min, in this embodiment, s =300, based on the obtained data information corresponding to each monitoring parameter, a monitoring matrix is established:
where H is the monitoring matrix, H 11 Data information corresponding to the 1 st monitoring parameter at the 1 st detection time h 1s Data information corresponding to the 1 st monitoring parameter at the s-th detection time h 21 Data information corresponding to the 2 nd monitoring parameter at the 1 st detection time h 2s Data information h corresponding to the 2 nd monitoring parameter at the s-th detection moment C1 Data information corresponding to the C-th monitoring parameter at the 1 st detection moment, h Cs And the data information corresponding to the C-th monitoring parameter at the s-th detection moment.
Each monitoring parameter corresponds to a row of data information in the monitoring matrix, and the monitoring matrix is subjected to normalization processing, so that the value of each data information is ensured to be in an interval (0, 1). According to the monitoring matrix, the data information corresponding to the same monitoring parameter at different detection moments in a time period is represented by each row of data information, and under the condition that the conveyor belt normally operates, the data information corresponding to the same monitoring parameter at different detection moments in the time period is approximately the same, so that the mutation point corresponding to each row of data information is obtained, specifically, the mutation point is obtained through a horizontal edge detection operator Sobel Level of Acquiring edge values corresponding to all data information in the monitoring matrix, recording the data information with the edge value of 1 as a mutation point, further acquiring a mutation point corresponding to each row of data information, and counting the number of the mutation points corresponding to each row of data information, wherein the mutation points do not exist in each row of data information, and if the mutation points do not exist in one row of data information, the number of the mutation points corresponding to the row of data information is 0.
The number of the mutation points can comprehensively reflect the overall abnormal condition of the corresponding row data information, and when the number of the mutation points is very small, the mutation points are considered to be isolated noise point data existing in the data information acquisition process of the monitoring parameters, namely the mutation points are considered not to be formed due to the abnormal condition of the corresponding monitoring parameters in the operation process of the conveyor belt; when the number of mutation points is large, it is considered that there is a high possibility that abnormality exists in the corresponding monitored parameter, and therefore the number of mutation points is compared with the set number.
And when the number of the catastrophe points corresponding to any row of data information is larger than the set number, calculating the abnormal evaluation value corresponding to the row of data information according to the maximum value, the minimum value and the variance corresponding to the row of data information. The value of the set number is 5, and an implementer can adjust the value of the set number according to the actual situation.
The abnormality evaluation value is:
wherein z is c An abnormal evaluation value h corresponding to the c-th line data information c,max Is the maximum value corresponding to the c-th row of data information, h c,min Is the minimum value, sigma, corresponding to the c-th row of data information c The variance corresponding to the information of the c-th row of data; exp (-) is an exponential function with a natural constant e as the base.
Abnormal evaluation value z c The larger the data information, the higher the possibility that the data information of the c-th row is abnormal. Sigma c (h c,max -h c,min ) Representing the fluctuation degree h of the data information of the c-th row c,max -h c,min Representing the difference value between the maximum value and the minimum value corresponding to the c-th row of data information, wherein the larger the difference value is, the higher the fluctuation degree of the c-th row of data information is represented, and sigma c The larger the size, the higher the fluctuation degree of the data information representing the c-th row. Since data information corresponding to the same monitoring parameter at different detection times within a time period should be substantially the same in the case where the conveyor belt is operating normally, the abnormality evaluation value z c The larger the data information, the higher the possibility that the data information of the c-th row is abnormal.
And 6, calculating the running state index of the conveyor belt according to the evaluation index corresponding to each real crack and the abnormal evaluation value, and monitoring the conveyor belt based on the running state index.
The running state index is the sum of all abnormal evaluation values and the sum of the evaluation indexes corresponding to all real cracks, and the formula of the running state index is expressed as follows:
wherein,
as an index of the operating state, P m The evaluation index corresponding to the mth real crack, M is the number of the real cracks, z c And C is the abnormal evaluation value corresponding to the C-th row of data information, and C is the number of rows in the monitoring matrix.
The operation state index represents the possibility of abnormity of the conveyor belt in the operation process, so the larger the operation state index is, the more possible the conveyor belt is to be abnormal. The more the number of the real cracks is, the worse the running state of the conveyor belt is; the larger the running state index is; the larger the evaluation index corresponding to the real crack is, the more obvious the crack state representing the real crack is, and the worse the running state of the conveyor belt is, the larger the running state index is. The larger the abnormality evaluation value is, the more likely the abnormality of the corresponding monitoring parameter is, and accordingly, the more likely the abnormality of the conveyor belt is, that is, the larger the running state index is.
It should be noted that, when the number of abrupt change points corresponding to any row of data information is less than or equal to the set number, the anomaly evaluation value corresponding to the row of data information is 0.
Further, the operation state indexes are normalized, so that the values of the operation state indexes are in intervals (0, 1), the operation state indexes are compared with the safe operation threshold value, when the operation state indexes are larger than the safe operation threshold value, the operation state of the conveyor belt is considered to be poor at the moment, an early warning prompt is sent out, a monitoring management center is informed timely, related personnel are prompted to overhaul the conveyor belt, and the occurrence of dangerous accidents caused by the fact that the operation state of the conveyor belt is too poor in the working process is avoided; when the running state index is less than or equal to the safe running threshold, the running state of the conveyor belt is considered to be good at the moment, and no early warning prompt is sent out; so far, the task of monitoring the conveyor belt is completed through the running state index.
The above-mentioned embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; the modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application, and are included in the protection scope of the present application.
Claims (8)
1. A method for monitoring the safe operation of an agricultural product transport conveyor, the method comprising the steps of:
acquiring a surface image of a conveyor belt, and preprocessing the surface image of the conveyor belt to obtain a gray image corresponding to the conveyor belt; filtering the gray level image in multiple directions by using filtering kernels with different scales respectively to obtain multiple feature extraction images; fusing a plurality of feature extraction images to obtain a saliency map;
extracting a straight line segment area in the saliency map, marking the straight line segment area as a suspected crack area, and obtaining a gradient distribution characteristic corresponding to each suspected crack area according to a gradient amplitude and a gradient direction angle corresponding to each suspected crack pixel point in the suspected crack area;
obtaining random indexes corresponding to the suspected crack areas according to the width and the length of the minimum external rectangle corresponding to each suspected crack area and the distance from all the suspected crack pixel points to the edge line of the conveyor belt;
recording the gray average value of all suspected crack pixel points corresponding to each suspected crack area as an intensity response value, obtaining an evaluation index corresponding to each suspected crack area based on the gradient distribution characteristics, the random index and the intensity response value corresponding to each suspected crack area, and recording the suspected crack area corresponding to the evaluation index larger than a set threshold value as a real crack;
collecting data information corresponding to different monitoring parameters of a conveyor belt in the operation process, constructing a monitoring matrix based on the data information, and acquiring mutation points corresponding to each row of data information;
and calculating the running state index of the conveyor belt according to the evaluation index corresponding to each real crack and the abnormal evaluation value, and monitoring the conveyor belt based on the running state index.
2. The method for monitoring the safe operation of the agricultural product transportation conveyor belt according to claim 1, wherein the method for fusing the plurality of feature extraction maps to obtain the saliency map specifically comprises the following steps: firstly, carrying out position alignment operation on feature extraction graphs corresponding to different directions under the same scale, and calculating a pixel mean value corresponding to the position according to pixel values of pixel points corresponding to the same position to obtain a fusion feature extraction graph corresponding to the scale; and then respectively carrying out significance processing on the fusion feature images corresponding to the non-scale scales to obtain fusion significance images corresponding to different scales, further distributing weights to each fusion significance image, and carrying out weighting processing on all fusion significance images according to the weights to obtain the significance images.
3. The method of claim 1, wherein said monitoring of said agricultural product transport conveyor belt is performed by a computer,
the method for acquiring the gradient distribution characteristics comprises the following steps: recording the gradient amplitude and the gradient direction angle corresponding to each suspected crack pixel point as gradient information, forming a gradient information set corresponding to each suspected crack region by the gradient information corresponding to all the suspected crack pixel points corresponding to each suspected crack region, recording the gradient information with the same value as the same gradient information, calculating the ratio of the frequency of each gradient information appearing in the gradient information set to the quantity of all the gradient information in the gradient information set, and determining the gradient distribution characteristics corresponding to each suspected crack region according to the variance value of the ratio and all the gradient direction angles in the gradient information set.
4. The agricultural product transportation conveyer belt safety operation monitoring method according to claim 1, wherein the random index obtaining method comprises the following steps:
and calculating the cumulative sum of the distances from all the suspected crack pixel points in each suspected crack area to the edge line of the conveyor belt, calculating the ratio of the width to the length of the minimum circumscribed rectangle, and determining a random index according to the cumulative sum and the ratio.
5. The method for monitoring the safe operation of the agricultural product transportation conveyor belt according to claim 1, wherein the evaluation index is:
wherein, P k Ran is an evaluation index corresponding to a suspected crack region k k Is a random index corresponding to the suspected crack area k,
the intensity response value corresponding to the suspected crack area k is obtained; gray k ' is a value after normalization of gradient distribution characteristics corresponding to a suspected fracture region k, theta k Forming an included angle between a straight line corresponding to the suspected crack area k and the edge line of the conveyor belt; e is a natural constant.
6. The method for monitoring the safe operation of the agricultural product transportation conveyor belt according to claim 1, wherein the method for acquiring the mutation point corresponding to each row of data information comprises the following steps: and acquiring edge values corresponding to all data information in the monitoring matrix through a horizontal edge detection operator, and recording the data information with the edge value of 1 as a mutation point so as to obtain the mutation point corresponding to each row of data information.
7. The method of monitoring the safe operation of a agricultural product transport conveyor belt according to claim 1, wherein the anomaly assessment value is:
wherein z is c An abnormal evaluation value h corresponding to the c-th line data information c,max Is the maximum value corresponding to the c-th row of data information, h c,min Is the minimum value, sigma, corresponding to the c-th row of data information c The variance corresponding to the c-th row of data information; exp (. Cndot.) is an exponential function based on a natural constant e.
8. The agricultural product transportation conveyer belt safe operation monitoring method according to claim 1, wherein the operation state index is a sum of an accumulation sum of evaluation indexes corresponding to all real cracks and a sum of all abnormal evaluation values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211281903.2A CN115684174A (en) | 2022-10-19 | 2022-10-19 | Agricultural product transportation conveyor belt safe operation monitoring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211281903.2A CN115684174A (en) | 2022-10-19 | 2022-10-19 | Agricultural product transportation conveyor belt safe operation monitoring method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115684174A true CN115684174A (en) | 2023-02-03 |
Family
ID=85066455
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211281903.2A Pending CN115684174A (en) | 2022-10-19 | 2022-10-19 | Agricultural product transportation conveyor belt safe operation monitoring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115684174A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116012384A (en) * | 2023-03-28 | 2023-04-25 | 梁山水泊胶带股份有限公司 | Method for detecting surface defects of whole-core flame-retardant conveying belt |
| CN116012383A (en) * | 2023-03-28 | 2023-04-25 | 山东鑫晟生物技术股份有限公司 | Data processing method for chondroitin sulfate production monitoring |
| CN116309565A (en) * | 2023-05-17 | 2023-06-23 | 山东晨光胶带有限公司 | High-strength conveyor belt deviation detection method based on computer vision |
| CN117557873A (en) * | 2024-01-12 | 2024-02-13 | 四川高速公路建设开发集团有限公司 | Tunnel face crack identification method based on image identification |
| CN118379639A (en) * | 2024-06-26 | 2024-07-23 | 大连乾曜科技有限公司 | Construction engineering quality detection method |
-
2022
- 2022-10-19 CN CN202211281903.2A patent/CN115684174A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116012384A (en) * | 2023-03-28 | 2023-04-25 | 梁山水泊胶带股份有限公司 | Method for detecting surface defects of whole-core flame-retardant conveying belt |
| CN116012383A (en) * | 2023-03-28 | 2023-04-25 | 山东鑫晟生物技术股份有限公司 | Data processing method for chondroitin sulfate production monitoring |
| CN116012383B (en) * | 2023-03-28 | 2023-06-09 | 山东鑫晟生物技术股份有限公司 | Data processing method for chondroitin sulfate production monitoring |
| CN116309565A (en) * | 2023-05-17 | 2023-06-23 | 山东晨光胶带有限公司 | High-strength conveyor belt deviation detection method based on computer vision |
| CN117557873A (en) * | 2024-01-12 | 2024-02-13 | 四川高速公路建设开发集团有限公司 | Tunnel face crack identification method based on image identification |
| CN117557873B (en) * | 2024-01-12 | 2024-04-05 | 四川高速公路建设开发集团有限公司 | Tunnel face crack identification method based on image identification |
| CN118379639A (en) * | 2024-06-26 | 2024-07-23 | 大连乾曜科技有限公司 | Construction engineering quality detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115684174A (en) | Agricultural product transportation conveyor belt safe operation monitoring method | |
| Zhang et al. | An efficient and reliable coarse-to-fine approach for asphalt pavement crack detection | |
| CN115239735B (en) | Communication cabinet surface defect detection method based on computer vision | |
| WO2022110804A1 (en) | Image noise measurement method and device based on local statistical information | |
| CN106296974B (en) | A kind of stickup paper money detection method and device | |
| CN107437243B (en) | Tire impurity detection method and device based on X-ray image | |
| CN109444169B (en) | Bearing defect detection method and system | |
| CN111310645A (en) | Overflow bin early warning method, device, equipment and storage medium for cargo accumulation amount | |
| US20180301025A1 (en) | Method and apparatus for predicting road conditions based on big data | |
| CN115861294B (en) | Concrete production abnormality detection method and device based on computer vision | |
| CN117876971B (en) | Building construction safety monitoring and early warning method based on machine vision | |
| CN111275705A (en) | Intelligent cloth inspecting method and device, electronic equipment and storage medium | |
| CN114332026A (en) | Visual detection method and device for scratch defects on surface of nameplate | |
| Zhao et al. | Image-based comprehensive maintenance and inspection method for bridges using deep learning | |
| CN111161233A (en) | Method and system for detecting defects of punched leather | |
| CN107436304A (en) | A kind of surface of concrete structure detection means | |
| CN116524205A (en) | Sewage aeration automatic detection and identification method | |
| CN114612403B (en) | Intelligent detection method and system for breakage defect of feeding belt | |
| CN107977531A (en) | A kind of method that ground resistance hard measurement is carried out based on image procossing and field mathematical model | |
| CN110992323A (en) | Steel strip detection method and device based on image processing and steel clamping detection method | |
| CN113989286A (en) | Deep learning method, device, equipment and storage medium for belt tearing strength | |
| CN112950594A (en) | Method and device for detecting surface defects of product and storage medium | |
| CN116309447A (en) | Dam slope crack detection method based on deep learning | |
| CN105759605A (en) | Nonlinear system defect detection and positioning algorithm based on adaptive parameter model particle filter (PF) | |
| Kembhavi et al. | Low-cost image processing system for evaluating pavement surface distress |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |