CN111754466B - Intelligent detection method for damage condition of conveyor belt - Google Patents
Intelligent detection method for damage condition of conveyor belt Download PDFInfo
- Publication number
- CN111754466B CN111754466B CN202010512893.3A CN202010512893A CN111754466B CN 111754466 B CN111754466 B CN 111754466B CN 202010512893 A CN202010512893 A CN 202010512893A CN 111754466 B CN111754466 B CN 111754466B
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- belt
- value
- target
- 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.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 56
- 239000011159 matrix material Substances 0.000 claims abstract description 33
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000005316 response function Methods 0.000 claims abstract description 14
- 230000001629 suppression Effects 0.000 claims abstract description 5
- 238000009499 grossing Methods 0.000 claims abstract description 4
- 238000012360 testing method Methods 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 15
- 238000001914 filtration Methods 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 7
- 230000000007 visual effect Effects 0.000 claims description 5
- 230000007797 corrosion Effects 0.000 claims description 4
- 238000005260 corrosion Methods 0.000 claims description 4
- 239000004973 liquid crystal related substance Substances 0.000 claims description 4
- 230000000877 morphologic effect Effects 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 239000012535 impurity Substances 0.000 claims description 3
- 238000004891 communication Methods 0.000 claims description 2
- 230000006835 compression Effects 0.000 claims description 2
- 238000007906 compression Methods 0.000 claims description 2
- 238000012217 deletion Methods 0.000 claims description 2
- 230000037430 deletion Effects 0.000 claims description 2
- 239000002245 particle Substances 0.000 claims description 2
- 238000012935 Averaging Methods 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 5
- 239000003245 coal Substances 0.000 abstract description 2
- 230000009347 mechanical transmission Effects 0.000 abstract description 2
- 238000001035 drying Methods 0.000 abstract 1
- 238000005272 metallurgy Methods 0.000 abstract 1
- 238000005065 mining Methods 0.000 abstract 1
- 230000007547 defect Effects 0.000 description 5
- 230000009466 transformation Effects 0.000 description 4
- 238000007689 inspection Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000011089 mechanical engineering Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/0008—Industrial image inspection checking presence/absence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration by the use of local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20132—Image cropping
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
-
- 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
Abstract
The invention discloses an intelligent detection method for damage conditions of a belt of a conveyor, which mainly solves the problems that the prior art cannot detect tearing damage conditions of the belt conveyor in real time in the running process and has high false detection rate. The implementation scheme is as follows: 1) Collecting an image of a belt to be tested; 2) Cutting, drying, normalizing, expanding and corroding, convolution operation and smoothing are sequentially carried out on the belt image to be detected, so as to obtain a new matrix M 1 The method comprises the steps of carrying out a first treatment on the surface of the 3) From a new matrix M 1 Calculating a corner response function corresponding to each pixel, performing non-maximum suppression arrangement on the corner response function, and solving a slope extremum difference S; 4) Comparing the slope extremum difference S with a given discrimination threshold k: if S is greater than k, the belt is torn; if S is less than k, the belt is not torn. The invention has the advantages of high detection speed and low false detection rate, can detect the damage condition of the belt in real time, and can be used for detecting the damage condition of the belt in the mechanical transmission process in mining industry, metallurgy and coal factories.
Description
Technical Field
The invention belongs to the technical field of detection instruments, and particularly relates to an intelligent detection device which can be used for detecting belt damage conditions in a mechanical transmission process.
Background
In many metallurgical, high-rise and coal industry plants, belt transport is certainly a more common and heavily used way of transporting goods, and the long transport distance and the heavy weight of the carried goods expose the belt to a greater risk of breakage and tearing. Once the conveyer belt is torn, a huge amount of materials are damaged, and serious economic loss is caused, so that safety detection and related maintenance on the damaged belt are needed. The traditional damage condition detection method is too single and mechanized, and has obvious defects and careless mistakes in the aspects of cost, detection time, detection accuracy, real-time performance, adjustability and the like. At present, domestic belt detection and monitoring systems are mainly divided into the following three types:
the first type is to use manual inspection in long-distance and large-transportation systems, so that the labor intensity is high, and the problem of missing inspection exists;
the second type is that the control part of the detection system is mostly in a man-machine interaction mode of artificial control, so that overload operation of the whole system is easy to occur, and potential safety hazards are brought;
the third type is to use a mechanical contact type detection device, such as a magnetic sensor, a pressure testing device and the like, and one of the most fatal defects of the devices is that the detection of the devices is carried out after an accident, the operation condition of a belt cannot be prejudged in advance, the real-time performance is lacking, the maintenance task of detection equipment is heavy, related parameter components are difficult to replace, and the work of an maintainer is difficult.
Based on the prior art, the target image is acquired through a high-speed industrial camera based on computer vision to finish digital conversion of a real target, and an effective detection algorithm is adopted to detect information in the image, so that actual belt running conditions and trends are reflected, the work of the target function to be achieved by the whole device is realized, and a plurality of people make attempts in China. Cheng Yue et al in the paper "visual detection of belt tear" ("mechanical engineering and Automation" 2018, 3) propose a visual auxiliary detection method to determine whether a belt has a crack or tear, so that the system gives an alarm or stops running. The idea of the party is as follows: in order to obtain high-quality belt working surface images, a special illumination light source is used for illuminating the belt working surface, so that light rays in the field of view of the camera are uniform. After the intelligent camera collects the belt image, the surface condition of the belt is identified through image processing, feature extraction and BP network judgment. When a suspected crack is found, an alarm is sent out timely to remind a worker to carry out further confirmation processing; when the belt tearing phenomenon is found, the working power supply of the belt conveyor should be cut off while alarming, so that further expansion of accidents is avoided. Here, the authors used a Canny edge detection operator to detect the edge of a belt crack, sent the detection result to a BP neural network, and analyzed the crack characteristics by the network to determine the damage condition of the belt. The method adopts visual assistance and has intellectualization. However, this method still has two disadvantages: firstly, the BP network has too high dependence on the acquired image, and if uneven light source irradiation or shake of the whole system occurs in the image acquisition process, adverse effect can be brought to the evaluation of the network; secondly, when the belt running speed is too high or too low, the detection accuracy is not very high.
Chen Yongbo et al patent "image processing-based belt tearing detection device" (bulletin No. CN 205151047U, bulletin day 2016.04.13, application day 2015.08.26) discloses an image processing-based belt tearing detection device, which collects complete belt picture data by adopting an area camera and a rotary encoder, an image processing platform accesses the area camera through an IP address to read pictures, performs a series of analyses such as preprocessing, defect detection, feature calculation and the like, and finally judges whether the belt has tearing or other defects. The device is convenient to maintain, and the problems of high false detection rate and the like of the traditional sensor detection are avoided, but the complexity of the system structure is certainly increased if the camera is required to scan to be completely matched with the belt running speed; meanwhile, the device has no special outstanding optimization on the detection algorithm, so that the improvement of the detection speed and the detection precision is influenced.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides an intelligent detection method for the damage condition of a conveyor belt, so as to realize real-time monitoring and detection of the damage condition of the belt in operation and improve the detection speed and precision.
In order to achieve the above object, the test method of the present invention comprises the steps of:
1) Collecting an image of a belt to be tested;
2) Processing the belt image to be detected to obtain detection parameters;
2a) Image data is read in accordance with the origin coordinates (x 0 ,y 0 ) The image is cropped with a width w and a height h to obtain a target image to be detected, wherein the origin coordinates (x 0 ,y 0 ) The width w and the height h are determined by the detection target range;
2b) Performing Gaussian filtering denoising on the target picture;
2c) Setting a limit threshold value with a maximum pixel value of 255 and a minimum pixel value of 50 of the denoised target picture, extracting a green component of the image, converting the image into a gray image, carrying out normalization processing on the green component, and enhancing the contrast ratio of the target and the background;
2d) The threshold T of binarization of the normalized image is obtained in a self-adaptive mode by adopting an Ojin method, and the threshold T is converted into a binary image;
2e) Performing expansion and corrosion operation on the binary image to eliminate miscellaneous points and interference areas in the picture;
2f) The small-area communication area deleting is carried out on the image with the impurity points and the interference areas eliminated, so that only detected lines exist in the image;
2g) Morphological refinement is carried out on the lines of the image after deletion, and single pixel point lines are extracted;
2h) Filtering each pixel of the single pixel point line by using a horizontal and vertical difference operator to obtain a Gaussian function in the single pixel point lineDerivative I in the horizontal and vertical directions of the image x And I y Calculating a matrix M for performing convolution operation on an original image;
2i) Gaussian smoothing filtering is performed on four elements of the matrix M to eliminate unnecessary isolated points and bumps, thereby obtaining a new matrix M 1 ；
2j) Using a new matrix M 1 Calculating a corner response function R corresponding to each pixel:
R＝detM 1 -α(traceM 1 ) 2 ，
wherein detM 1 ＝λ 1 λ 2 ，traceM 1 ＝λ 1 +λ 2 ，λ 1 、λ 2 Respectively a new matrix M 1 The transformation degree of matrix vectors, namely the transformation degree of straight lines at two sides of the corner points, wherein alpha is a corner point threshold value and the value is 0.04-0.06;
2k) Local non-maximum suppression is carried out on the corner response function R, namely, the corner which is not the maximum is suppressed, and the local maximum R is obtained max In the corner response function matrix R, if each element R (i, j) in the corner response function matrix R is a local maximum R in the vicinity 8 o' clock max The point of the position coordinate is then considered to be a corner point;
2 l) arranging the R values of the detected corner points, selecting corner point positions corresponding to N sequentially largest values in the R value sequence, calculating slope values between any two points of the coordinates of the N corner point positions, and calculating the difference between the maximum value and the minimum value, namely slope extremum difference S;
3) Judging whether the belt is torn or not according to the detection parameters, namely comparing the slope extremum difference S with a given judging threshold k:
if the slope extremum difference S is greater than the discrimination threshold k, the belt is torn;
if the slope extremum difference S is less than the discrimination threshold k, the belt is not torn.
The invention has the following advantages:
first: the detection method is simple to realize and the detection result is high in accuracy.
Most of detection means in the current belt detection field are common alarm signal detection or manual overhaul, and the implementation process is time-consuming and complex; the detection method is simple to realize, and can rapidly detect the condition of the belt in operation in real time; most of the detection methods adopted at present in China are single and lack of precision, the detection method used by the invention has intelligence, the accurate detection algorithm is utilized to detect the picture processed by the system, and the detection result is accurate.
Second,: the detection implementation means is novel, and the detection method has good packagability.
The method utilizes a laser generator to emit green laser beams so as to uniformly irradiate the green laser beams on the belt, and a high-speed industrial camera is loaded at a position convenient for shooting the laser beams, so that the acquired images are ensured to be uniform and clear.
Drawings
FIG. 1 is a flow chart of an implementation of the present invention for detecting a damaged condition of a belt conveyor.
Detailed Description
The following detailed description of the invention is made with reference to the accompanying drawings:
referring to fig. 1, the method for detecting the damage condition of the conveyor belt according to this example is implemented as follows:
and step 1, acquiring an image of the belt to be detected.
1.1 Mounting test device:
loading a linear laser generator, adjusting the linear laser generator to a position to ensure that the emitted green laser beam can irradiate the side surface of the belt, and connecting a control circuit of the linear laser generator with a test box;
loading a high-speed industrial camera, and adjusting the position of the high-speed industrial camera to ensure that a detection target image with an optimal visual angle can be shot, and simultaneously connecting the high-speed industrial camera with a test box;
the test box sends a control instruction to enable the linear laser generator to generate linear green laser beams, the intensity of the linear green laser beams is based on the best resolution of the high-speed industrial camera, and the linear green laser beams transversely and uniformly irradiate on the surface of the belt to be tested;
1.2 Collecting an image of a belt to be measured:
the test box controls the high-speed industrial camera to start shooting the belt to be tested in real time, namely, a control module of the test box acquires continuous image frames and inputs the continuous image frames into an image processing module of the test box, and the image processing module of the test box converts data transmitted by the high-speed industrial camera into RGB images to complete acquisition of the belt images to be tested.
And 2, processing the belt image to be detected to obtain detection parameters.
2.1 Reading in image data according to the origin coordinates (x) of the target region 0 ,y 0 ) The image is cropped with a width w and a height h to obtain a target image to be detected, wherein the origin coordinates (x 0 ,y 0 ) The width w and the height h are determined by the detection target range;
2.2 Weighted average is carried out on the whole image, and Gaussian filtering denoising of the target picture is realized: the value of each pixel point is obtained by weighted average of the pixel point and other pixel values in the neighborhood, namely, each pixel in the template scanning image is used, and the weighted average gray value of the pixels in the neighborhood determined by the template is used for replacing the value of the central pixel point of the template, wherein the formula is as follows:
wherein G (x, y) is an element in a 3×3 mask matrix, sigma is a Gaussian filter mask of 0.5, and x, y is the relative position coordinates of the 3×3 mask;
2.3 Setting a limit threshold value with a maximum pixel value of 255 and a minimum pixel value of 50 of the denoised target picture, extracting a green component of the image, converting the green component into a gray image, carrying out normalization processing on the green component, and enhancing the contrast ratio of the target and the background; image normalization is performed by the following formula:
y＝(x-min value )/(max value -min value )，
wherein, x and y are values before and after normalization, max value 、min value Respectively the maximum value and the minimum value of the gray scale of the original image;
2.4 Self-adaptively acquiring a normalized image binarization threshold T by adopting an Ojin method, and converting the normalized image binarization threshold T into a binary image:
2.4.1 Dividing the pixel values of the picture into [0,1,2, …,255 ]]Level of n i Number of pixels, n, representing each horizontal pixel value L The number of the maximum horizontal image values of the picture is the total number of the pixels:
N＝n 1 +n 2 +…+n i +…+n L ；
2.4.2 Setting the segmentation threshold value of the target and the background as T, and recording the number of pixels with the pixel gray value smaller than the threshold value T in the image as N 0 The number of pixels with the pixel gray value larger than the threshold value T is recorded as N 1 Calculating the proportion omega of the pixel number of the target to the whole image 0 And its average gray scale mu 0 Calculating the proportion omega of the number of background pixel points to the whole image 1 And its average gray scale mu 1 ：
ω 0 ＝N 0 /N，ω 1 ＝N 1 /N，
Wherein, the value of T is one integer value of 0 to 255;
2.4.3 Calculating the inter-class variance g) from the parameters obtained in 2.4.2):
g＝ω 0 ω 1 (μ 0 -μ 1 ) 2 ；
2.4.4 Obtaining a segmentation threshold T which maximizes the inter-class variance g by adopting a traversing method;
2.5 Corrosion and expansion operations are carried out on the binary image to eliminate miscellaneous points and interference areas in the picture:
the etching operation is to shrink the image boundary to make the target area scope smaller so as to eliminate small and nonsensical targets, and the formula is as follows:
wherein A is the target image, B is a 2×2 full-matrix template, and A.sup.B is defined as follows: corroding A by using a structure B, when the origin of B is translated to a pixel (x, y) of an image A, if B is completely contained in an overlapped area of the image A at the position of (x, y), assigning 1 to the pixel (x, y) corresponding to the output image, otherwise, assigning 0 to the pixel (x, y);
the expansion operation is to incorporate the background point contacted by the target area into the target object, so that the target boundary is expanded outwards, the range of the target area is enlarged, and certain holes in the target area are filled and small particle noise contained in the target area is eliminated, and the formula is as follows:
wherein A is a target image, B is a 2 x 2 full-matrix structure template,is defined as follows: expanding A by using a structure B, translating an origin of a structural element B to an image pixel (x, y) position, if the intersection of B and A at the image pixel (x, y) is not null, assigning 1 to the pixel (x, y) corresponding to the output image, otherwise, assigning 0;
2.6 The small-area connected region is deleted from the image after the miscellaneous points and the interference regions are eliminated, so that only detected lines exist in the image, namely, the outline of the small-area connected region in the image is obtained firstly, then the area is calculated according to the outline, and the pixel values of all connected regions with the area smaller than a threshold value H are set to 0 to become a background, wherein the value of H is 50-80;
2.7 Morphological refinement is carried out on the lines of the deleted image, and single pixel point lines are extracted; the operation is that the lines with the line width larger than 1 pixel in the original image are thinned to be only one pixel wide, so as to form a framework, namely, the lines are deprived layer by layer from the edges of the lines until one pixel is left in the lines, so that the compression of the data amount of the original image is realized, and the basic topological structure of the shape of the original image is kept unchanged;
2.8 Filtering each pixel of the single pixel point line by utilizing a horizontal and vertical difference operator to obtain derivatives I of the Gaussian function in the horizontal direction and the vertical direction of the image x And I y A matrix M for performing convolution operation on the original image is calculated:
wherein I is x And I y Derivatives of the gaussian function in the horizontal x and vertical y directions of the image, respectively;
2.9 Gaussian smoothing filtering is performed on four elements of the matrix M to eliminate unnecessary isolated points and bumps to obtain a new matrix M 1 ：
Wherein, the liquid crystal display device comprises a liquid crystal display device,is a window function, (x, y) is the corresponding pixel coordinate position in the window, the window size reference is set to 3 x 3, and the sigma value is 0.5;
2.10 Using a new matrix M 1 Calculating a corner response function R corresponding to each pixel:
R＝detM 1 -α(traceM 1 ) 2 ，
wherein detM 1 ＝λ 1 λ 2 ，traceM 1 ＝λ 1 +λ 2 ，λ 1 、λ 2 Respectively a new matrix M 1 Is used to determine the characteristic value of the (c) for the (c),the transformation degree of the matrix vector, namely the transformation degree of the straight lines at two sides of the corner point is represented; alpha is a corner threshold value and has a value of 0.04-0.06;
2.11 Local non-maximum suppression of the angular point response function R, i.e. suppression of angular points other than maximum, obtaining a local maximum R max In the corner response function matrix R, if each element R (i, j) in the corner response function matrix R is a local maximum R in the vicinity 8 o' clock max The point of the position coordinate is then considered to be a corner point;
2.12 Arranging the R values of the detected corner points, selecting corner point positions corresponding to N sequentially largest values in the R value sequence, solving the slope values between any two points of the coordinates of the N corner point positions, and solving the difference between the maximum value and the minimum value of the slope values, namely the slope extremum difference S;
2.13 Judging whether the belt is torn or not according to the detection parameters, namely comparing the slope extremum difference S with a given judging threshold k, wherein the judging threshold is an experimental experience value, and the value range of the judging threshold is 0.3-0.5 through experiments:
if the slope extremum difference S is greater than the discrimination threshold k, the belt is torn;
if the slope extremum difference S is less than the discrimination threshold k, the belt is not torn.
The foregoing is a specific example of the invention and is not intended to limit the invention in any way, and it will be apparent that modifications in different forms and parameters may be made without departing from the spirit and principles of the invention, but are within the scope of the invention.
Claims (10)
1. A method of detecting a damaged condition of a conveyor belt, comprising:
1) Collecting an image of a belt to be tested;
2) Processing the belt image to be detected to obtain detection parameters;
2a) Image data is read in accordance with the origin coordinates (x 0 ,y 0 ) The image is cropped with a width w and a height h to obtain a target image to be detected, wherein the origin coordinates (x 0 ,y 0 ) The width w and the height h are determined by the detection target range;
2b) Performing Gaussian filtering denoising on the target picture;
2c) Setting a limit threshold value with a maximum pixel value of 255 and a minimum pixel value of 50 of the denoised target picture, extracting a green component of the image, converting the image into a gray image, carrying out normalization processing on the green component, and enhancing the contrast ratio of the target and the background;
2d) The threshold T of binarization of the normalized image is obtained in a self-adaptive mode by adopting an Ojin method, and the threshold T is converted into a binary image;
2e) Performing expansion and corrosion operation on the binary image to eliminate miscellaneous points and interference areas in the picture;
2f) The small-area communication area deleting is carried out on the image with the impurity points and the interference areas eliminated, so that only detected lines exist in the image;
2g) Morphological refinement is carried out on the lines of the image after deletion, and single pixel point lines are extracted;
2h) Filtering each pixel of the single pixel point line by using a horizontal and vertical difference operator to obtain derivatives I of the Gaussian function in the horizontal direction and the vertical direction of the image x And I y Calculating a matrix M for performing convolution operation on an original image;
2i) Gaussian smoothing filtering is performed on four elements of the matrix M to eliminate unnecessary isolated points and bumps, thereby obtaining a new matrix M 1 ；
2j) Using a new matrix M 1 Calculating a corner response function R corresponding to each pixel:
R＝detM 1 -α(traceM 1 ) 2 ，
wherein detM 1 ＝λ 1 λ 2 ，traceM 1 ＝λ 1 +λ 2 ，λ 1 、λ 2 Respectively a new matrix M 1 Alpha is a corner threshold value, and the value is 0.04-0.06;
2k) Local non-maximum suppression is carried out on the corner response function R, namely, the corner which is not the maximum is suppressed, and the local maximum R is obtained max In the corner response function matrix RIf each element R (i, j) in the corner response function matrix R is a local maximum R in the vicinity 8 max The point of the position coordinate is then considered to be a corner point;
2 l) arranging the R values of the detected corner points, selecting corner point positions corresponding to N sequentially largest values in the R value sequence, calculating slope values between any two points of the coordinates of the N corner point positions, and calculating the difference between the maximum value and the minimum value, namely slope extremum difference S;
3) Judging whether the belt is torn or not according to the detection parameters, namely comparing the slope extremum difference S with a set judging threshold k:
if the slope extremum difference S is greater than the discrimination threshold k, the belt is torn;
if the slope extremum difference S is less than the discrimination threshold k, the belt is not torn.
2. The method of claim 1, wherein 1) capturing an image of the belt under test is accomplished by:
1a) Loading a linear laser generator, adjusting the linear laser generator to a position to ensure that the emitted green laser beam can irradiate the side surface of the belt, and connecting a control circuit of the linear laser generator with a test box;
1b) Loading a high-speed industrial camera, and adjusting the position of the high-speed industrial camera to ensure that a detection target image with an optimal visual angle can be shot, and simultaneously connecting the high-speed industrial camera with a test box;
1c) The test box sends a control instruction to enable the linear laser generator to generate linear green laser beams, the intensity of the linear green laser beams is based on the best resolution of the high-speed industrial camera, and the linear green laser beams transversely and uniformly irradiate on the surface of the belt to be tested;
1d) The test box controls the high-speed industrial camera to start shooting the belt to be tested in real time, namely, a control module of the test box acquires continuous image frames and inputs the continuous image frames into an image processing module of the test box, and the image processing module of the test box converts data transmitted by the high-speed industrial camera into RGB images to complete acquisition of the belt images to be tested.
3. The method of claim 1 wherein the denoising of the target image in 2 b) is a weighted average of the entire image, the value of each pixel being obtained by weighted averaging itself and other pixel values in the neighborhood, i.e., each pixel in the image is scanned with a template, and the value of the center pixel of the template is replaced with the weighted average gray value of the pixel in the neighborhood determined by the template, the formula is as follows:
where G (x, y) is an element in a 3×3 mask matrix, σ is a gaussian filter mask of 0.5, and x, y is the relative position coordinates of the 3×3 mask.
4. The method according to claim 1, characterized in that the normalization of the image in 2 c) is performed by:
y＝(x-min value )/(max value -min value )，
wherein, x and y are values before and after normalization, max value 、min value Respectively, the maximum value and the minimum value of the gray scale of the original image.
5. The method according to claim 1, wherein in 2 d), the normalized image binarization threshold T is adaptively obtained by using the oxford method, and is converted into a binary image, which is implemented as follows:
2d1) Dividing pixel values of a picture into [0,1,2, …,255 ]]Level of n i Number of pixels, n, representing each horizontal pixel value L The number of the maximum horizontal image values of the picture is the total number of the pixels:
N＝n 1 +n 2 +…+n i +…+n L ；
2d2) Let the segmentation threshold of the target and the background be T, and record the number of pixels with the pixel gray value smaller than the threshold T as N 0 Pixels with pixel gray values larger than the threshold value T are arrangedThe number is recorded as N 1 Calculating the proportion omega of the pixel number of the target to the whole image 0 And its average gray scale mu 0 Calculating the proportion omega of the number of background pixel points to the whole image 1 And its average gray scale mu 1 ：
ω 0 ＝N 0 /N，ω 1 ＝N 1 /N，
Wherein, the value of T is one integer value of 0 to 255;
2d3) Calculating the inter-class variance g according to the parameters obtained in 2d 2):
g＝ω 0 ω 1 (μ 0 -μ 1 ) 2 ，
2d4) The segmentation threshold T which maximizes the inter-class variance g is obtained by adopting a traversal method.
6. The method of claim 1, wherein the operation of corroding and expanding the binary image in 2 e) is performed as follows:
the corrosion to the binary image is to shrink the boundary of the image, so that the range of the target area is reduced to eliminate small and meaningless targets, and the formula is as follows:
wherein A is the target image, B is a 2×2 full-matrix template, and A.sup.B is defined as follows: corroding A by using a structure B, when the origin of B is translated to a pixel (x, y) of an image A, if B is completely contained in an overlapped area of the image A at the position of (x, y), assigning the pixel (x, y) corresponding to the output image to be 1, otherwise, assigning the pixel (x, y) to be 0;
the expansion operation is performed on the binary image, namely, the background points contacted with the target area are combined into the target object, so that the target boundary is expanded outwards, the range of the target area is enlarged, certain holes in the target area are filled, and small particle noise contained in the target area is eliminated, wherein the formula is as follows:
wherein A is a target image, B is a 2 x 2 full-matrix structure template,is defined as follows: and expanding A by using the structure B, translating the origin of the structural element B to the position of an image pixel (x, y), if the intersection of B and A at the position of the image pixel (x, y) is not null, assigning 1 to the pixel (x, y) corresponding to the output image, and otherwise, assigning 0 to the pixel (x, y).
7. The method of claim 1, wherein the small-area connected region is deleted from the image after removing the impurity points and the interference regions in 2 f), wherein the contour of the small-area connected region in the image is obtained first, then the area is calculated according to the contour, and the pixel values of all connected regions with the area smaller than a threshold value H are set to 0 and become the background, wherein the value of H is 50-80.
8. The method of claim 1 wherein 2 g) performs morphological refinement on the lines of the pruned image to extract single pixel lines, wherein the lines with the line width greater than 1 pixel in the original image are refined to be only one pixel wide to form a 'skeleton', i.e. the lines are stripped layer by layer from the edges of the lines until the lines remain one pixel, so as to achieve compression of the original image data amount and keep the basic topological structure of the shape unchanged.
9. The method according to claim 1, characterized in that the matrix M in 2 h) is represented as follows:
wherein I is x And I y The derivatives of the gaussian function in the horizontal x and vertical y directions of the image, respectively.
10. The method according to claim 1, characterized in that the new matrix M in 2 i) 1 The expression is as follows:
wherein, the liquid crystal display device comprises a liquid crystal display device,is a window function, (x, y) is the corresponding pixel coordinate location within the window, the window size reference is set to 3 x 3, and the sigma value is 0.5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010512893.3A CN111754466B (en) | 2020-06-08 | 2020-06-08 | Intelligent detection method for damage condition of conveyor belt |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010512893.3A CN111754466B (en) | 2020-06-08 | 2020-06-08 | Intelligent detection method for damage condition of conveyor belt |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111754466A CN111754466A (en) | 2020-10-09 |
| CN111754466B true CN111754466B (en) | 2023-07-28 |
Family
ID=72676457
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010512893.3A Active CN111754466B (en) | 2020-06-08 | 2020-06-08 | Intelligent detection method for damage condition of conveyor belt |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111754466B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112374081A (en) * | 2020-11-12 | 2021-02-19 | 泉州冰点科技有限公司 | Large-scale conveyer belt surface crack detection method and system |
| CN112598647A (en) * | 2020-12-24 | 2021-04-02 | 凌云光技术股份有限公司 | Method for detecting weak line defect under arc-shaped surface gradual change background |
| CN112884784A (en) * | 2021-03-11 | 2021-06-01 | 南通大学 | Image-based lens detection and front-back judgment method |
| CN113658136B (en) * | 2021-08-17 | 2023-10-13 | 燕山大学 | Deep learning-based conveyor belt defect detection method |
| CN114013955A (en) * | 2021-10-27 | 2022-02-08 | 北方魏家峁煤电有限责任公司 | Belt longitudinal tearing detection method and system based on infrared camera |
| CN113724258B (en) * | 2021-11-02 | 2022-02-08 | 山东中都机器有限公司 | Conveyor belt tearing detection method and system based on image processing |
| CN117495857B (en) * | 2023-12-29 | 2024-04-02 | 合肥金星智控科技股份有限公司 | Belt tearing edge detection method, system, equipment and medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109359594A (en) * | 2018-10-17 | 2019-02-19 | 中国矿业大学 | A kind of mine belt method for detecting abnormality and device based on video processing |
| WO2019134252A1 (en) * | 2018-01-03 | 2019-07-11 | 东南大学 | Method and device for automated portrayal and accurate measurement of width of structural crack |
| CN110111303A (en) * | 2019-04-04 | 2019-08-09 | 上海大学 | A kind of large-scale carrier strip tearing intelligent fault detection method based on dynamic image |
| CN110211107A (en) * | 2019-05-28 | 2019-09-06 | 太原理工大学 | A kind of mining adhesive tape damage detecting method based on dual-band infrared image |
| CN110844518A (en) * | 2019-12-19 | 2020-02-28 | 安徽理工大学 | Conveying belt tearing detection method based on FPGA |
-
2020
- 2020-06-08 CN CN202010512893.3A patent/CN111754466B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019134252A1 (en) * | 2018-01-03 | 2019-07-11 | 东南大学 | Method and device for automated portrayal and accurate measurement of width of structural crack |
| CN109359594A (en) * | 2018-10-17 | 2019-02-19 | 中国矿业大学 | A kind of mine belt method for detecting abnormality and device based on video processing |
| CN110111303A (en) * | 2019-04-04 | 2019-08-09 | 上海大学 | A kind of large-scale carrier strip tearing intelligent fault detection method based on dynamic image |
| CN110211107A (en) * | 2019-05-28 | 2019-09-06 | 太原理工大学 | A kind of mining adhesive tape damage detecting method based on dual-band infrared image |
| CN110844518A (en) * | 2019-12-19 | 2020-02-28 | 安徽理工大学 | Conveying belt tearing detection method based on FPGA |
Non-Patent Citations (1)
| Title |
|---|
| 像素自相关矩阵的阈值自适应角点检测算法;邓小炼;杜玉琪;王长耀;王晓花;;农业工程学报(第18期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111754466A (en) | 2020-10-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111754466B (en) | Intelligent detection method for damage condition of conveyor belt | |
| CN106683075B (en) | Method for detecting bolt defects at cross arm of power transmission line tower | |
| CN111260616A (en) | Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization | |
| CN110033431B (en) | Non-contact detection device and detection method for detecting corrosion area on surface of steel bridge | |
| CN111369516B (en) | Transformer bushing heating defect detection method based on infrared image recognition | |
| CN110443791B (en) | Workpiece detection method and device based on deep learning network | |
| CN111539927B (en) | Detection method of automobile plastic assembly fastening buckle missing detection device | |
| CN109166125A (en) | A kind of three dimensional depth image partitioning algorithm based on multiple edge syncretizing mechanism | |
| CN112419298B (en) | Bolt node plate rust detection method, device, equipment and storage medium | |
| CN113538503B (en) | Solar panel defect detection method based on infrared image | |
| CN112101138B (en) | Bridge inhaul cable surface defect real-time identification system and method based on deep learning | |
| CN106651893A (en) | Edge detection-based wall body crack identification method | |
| CN112528979B (en) | Transformer substation inspection robot obstacle distinguishing method and system | |
| CN112270658A (en) | Elevator steel wire rope detection method based on machine vision | |
| CN111899219A (en) | Image identification method and system for power transmission line machine patrol | |
| CN112508911A (en) | Rail joint touch net suspension support component crack detection system based on inspection robot and detection method thereof | |
| CN110276787B (en) | Conductor galloping monitoring method based on marker image detection | |
| Zhou et al. | UAV vision detection method for crane surface cracks based on Faster R-CNN and image segmentation | |
| CN112288682A (en) | Electric power equipment defect positioning method based on image registration | |
| CN113705564B (en) | Pointer type instrument identification reading method | |
| CN113971681A (en) | Edge detection method for belt conveyor in complex environment | |
| CN110223332B (en) | Bridge crack calibration method | |
| CN106530292A (en) | Strip steel surface defect image rapid identification method based on line scanning camera | |
| CN110852990A (en) | Rubber tree oblique cutter mark detection method based on image processing | |
| CN114266748B (en) | Method and device for judging surface integrity of process board in field of rail traffic overhaul |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |