CN114782330B - Grate abnormity detection method and system based on artificial intelligence - Google Patents
Grate abnormity detection method and system based on artificial intelligence Download PDFInfo
- Publication number
- CN114782330B CN114782330B CN202210346787.1A CN202210346787A CN114782330B CN 114782330 B CN114782330 B CN 114782330B CN 202210346787 A CN202210346787 A CN 202210346787A CN 114782330 B CN114782330 B CN 114782330B
- Authority
- CN
- China
- Prior art keywords
- grate
- fire grate
- points
- artificial intelligence
- fire
- 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 41
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 28
- 230000007547 defect Effects 0.000 claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 39
- 230000011218 segmentation Effects 0.000 claims abstract description 11
- 238000012216 screening Methods 0.000 claims abstract description 6
- 230000002159 abnormal effect Effects 0.000 claims description 21
- 238000009826 distribution Methods 0.000 claims description 20
- 230000005856 abnormality Effects 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 11
- 238000005286 illumination Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 abstract 1
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000001914 filtration Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 238000003723 Smelting Methods 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000003708 edge detection Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 239000010959 steel Substances 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005266 casting Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003245 coal Substances 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000007493 shaping process Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000009423 ventilation Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
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
-
- 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
- G06T7/13—Edge detection
-
- 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
Abstract
The invention relates to the technical field of artificial intelligence, in particular to a fire grate abnormity detection method and system based on artificial intelligence. The method comprises the following steps: collecting a surface image of the fire grate at a overlooking visual angle, and extracting the edge of the fire grate; carrying out Hough straight line detection on pixel points with the reliability higher than a preset threshold value to obtain a plurality of coordinate points mapped in Hough space by a plurality of straight lines; selecting a straight line segment with the longest length as a first grate bar, and acquiring a corresponding first coordinate point; screening out coordinate points with the same abscissa as the first coordinate points to serve as suspected grate points, and taking the mode of the distance between adjacent suspected grate points as a grate interval; searching according to the grate intervals by taking the first coordinate point as a center to obtain all grate points; and obtaining a semantic segmentation image of the surface image, extracting grate bars corresponding to the grate points from the semantic segmentation image, and performing reverse color on the grate bars and the edge of the grate to obtain a defect area. The embodiment of the invention can accurately and quickly extract the defect area in the grate.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to a fire grate abnormity detection method and system based on artificial intelligence.
Background
The boiler grate is a steel smelting structure, also is the boiler grate, and the boiler grate mainly is many grate bars and constitutes, and the centre is the space for increase the volume of throwing coal, increase the ventilation of boiler, improve the availability factor of boiler, further improve steel smelting efficiency.
The burr appears in the boiler grate gap department easily behind the casting shaping, can block up the space position between the grate bar, when the grate gap appears blockking up, will produce very big harm to the normal operating of boiler, will lead to when serious to be forced to shut down the stove, simultaneously, when the grate shape size that produces etc. appears unusually, will unable use in throwing into corresponding boiler, causes the waste of resource.
At present, the quality of the boiler grate is generally detected in a manual mode, but the manual detection efficiency is low, and the boiler grate is easily damaged in the detection process.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a fire grate abnormity detection method and system based on artificial intelligence, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides a method for detecting abnormal fire grates based on artificial intelligence, the method comprising the steps of:
collecting a surface image of a fire grate at a top view angle, and extracting the edge of the fire grate;
calculating the confidence coefficient of the gradient of each pixel point in the surface image, and performing Hough line detection on the pixel points with the confidence coefficient higher than a preset threshold value to obtain a plurality of coordinate points mapped in Hough space by a plurality of lines;
selecting a straight line segment with the longest length as a first grate bar, and acquiring a corresponding first coordinate point; screening out coordinate points with the same abscissa as the first coordinate points to serve as suspected grate points, and taking the mode of the distance between adjacent suspected grate points as a grate interval; searching according to the grate intervals by taking the first coordinate point as a center to obtain all grate points;
and obtaining a semantic segmentation image of the surface image, extracting grate bars corresponding to the grate points from the semantic segmentation image, and performing reverse color on the grate bars and the grate edges to obtain a defect area.
Preferably, the method further comprises the steps of:
acquiring the integral characteristics of the fire grate according to the edge of the fire grate; the integral features include dimensional and shape features;
acquiring a distribution index of the defect area; taking the area and the position information of the defect area and the distribution index as local characteristics of the fire grate;
and calculating the abnormal degree of the fire grate according to the overall characteristic and the local characteristic.
Preferably, the distribution index obtaining method includes:
clustering the central points of all the defect areas to obtain the number of clustering categories and the number of discrete points;
and calculating the distribution index according to the number of the cluster categories and the number of the discrete points.
Preferably, the distribution index obtaining method includes:
and acquiring the shortest distance between every two defect areas, and taking the sum of all the shortest distances as the distribution index.
Preferably, the step of obtaining the dimensional characteristics comprises:
and obtaining the detection radius of the fire grate according to the edge of the fire grate, and taking the difference value of the detection radius and the standard radius of a standard fire grate as the size characteristic.
Preferably, the step of obtaining the shape feature includes:
and counting the number of pixel points in the edge of the fire grate to serve as the area of the fire grate, calculating the circularity of the edge of the fire grate according to the area of the fire grate and the length of the edge of the fire grate, and taking the circularity as the shape characteristic.
Preferably, the confidence coefficient calculation process includes:
and calculating the gradient amplitude of each pixel by utilizing the gradient of each pixel point in the horizontal direction and the vertical direction, and calculating the confidence coefficient according to the gradient amplitude.
Preferably, the step of obtaining all fire grate points further comprises the steps of:
and counting the number of the fire grate points, comparing the number with the number of the fire grate bars of the standard fire grate, and judging whether the fire grate has redundant attachments or lacks the fire grate bars.
Preferably, the step of extracting the edge of the grate further comprises:
and removing the noise of the surface image to obtain a de-noised image, and then carrying out illumination equalization processing on the de-noised image.
In a second aspect, another embodiment of the present invention provides an artificial intelligence based fire grate abnormality detecting system, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the artificial intelligence based fire grate abnormality detecting method when executing the computer program.
The embodiment of the invention at least has the following beneficial effects:
firstly, extracting the edge of a fire grate from a surface image of the fire grate, then extracting a straight line in the surface image through optimized Hough straight line detection, finding a first fire grate bar according to the numerical value of each straight line segment in the Hough space, finding all the fire grate bars according to the properties that the fire grate bars are parallel to each other and the fire grate intervals are the same, and performing color reversal on the fire grate bars and the edge to obtain a defect area. The embodiment of the invention utilizes the image data to carry out the abnormity detection on the boiler grate based on an artificial intelligence mode, can effectively reduce the system calculation amount, improves the detection efficiency, has more accurate detection result and reduces the false detection rate.
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 description of the embodiments or 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 flowchart illustrating steps of a method for detecting abnormal conditions of a grate based on artificial intelligence according to an embodiment of the present invention.
Detailed Description
To further illustrate the technical means and effects of the present invention for achieving the predetermined objects, the following detailed description of the fire grate abnormality detection method and system based on artificial intelligence according to the present invention, the detailed implementation, structure, features and effects thereof will be provided in conjunction with the accompanying drawings and the 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.
The following describes a specific scheme of the grate abnormality detection method and system based on artificial intelligence in detail with reference to the accompanying drawings.
Referring to fig. 1, a flowchart illustrating steps of a method for detecting abnormal conditions of a grate based on artificial intelligence according to an embodiment of the present invention is shown, wherein the method comprises the following steps:
and S001, collecting the surface image of the fire grate at the overlooking view angle and extracting the edge of the fire grate.
The method comprises the following specific steps:
1. and collecting a surface image of the fire grate.
The method comprises the steps that for the round fire grates produced and processed in large scale, image collection is carried out when the round fire grates are conveyed through a conveying belt, a camera is arranged at the position, close to the tail portion, of the conveying belt for conveying the boiler fire grates and located right above the conveying belt, images on the surfaces of the boiler fire grates are collected in a overlooking angle, and the conveying speed of the conveying belt and the frame rate of images shot by the camera are set according to actual conditions.
2. The surface image is pre-processed.
Considering that the production and processing environment of boiler parts is complex, in order to avoid the influence of uneven illumination and noise, the collected surface image is subjected to primary processing, the noise of the surface image is removed, a de-noised image is obtained, and then illumination equalization processing is performed on the de-noised image.
Specifically, the image is filtered to obtain a denoised image. There are many image filtering algorithms, and preferably, the embodiment of the present invention processes an image by using an adaptive median filtering algorithm to eliminate a large amount of noise in the image. In other embodiments, an image filtering algorithm that can achieve the same effect, such as median filtering, mean filtering, gaussian filtering, and the like, may also be used.
The filtered image is subjected to illumination equalization processing, so that the influence of illumination unevenness on the extraction of the subsequent grate characteristics is avoided, the existing image illumination equalization methods are many, and preferably, the embodiment of the invention adopts histogram equalization to perform equalization processing on the de-noised image so as to eliminate the influence of the image illumination unevenness. In other embodiments, an illumination equalization method such as a gamma transform method, a gray scale transform method, and a thresholding method may be used to achieve the same effect.
3. And extracting the edge of the fire grate.
And carrying out gray processing on the preprocessed image, and extracting an edge image of the boiler grate by adopting an edge detection algorithm.
As an example, the embodiment of the invention performs edge detection on the gray-scale image by using a sobel operator to extract edge information of a boiler grate.
Further, acquiring the integral characteristics of the fire grate according to the edge of the fire grate; the overall characteristics include dimensional and shape characteristics.
3.1 obtaining the detection radius of the fire grate according to the edge of the fire grate, and taking the difference value of the detection radius and the standard radius of the standard fire grate as the size characteristic.
Δr＝|r-r 0 |
Where Δ r represents a dimensional characteristic, r represents a detection radius, r 0 Indicating a standard radius.
3.2, counting the number of pixel points in the edge of the fire grate to be used as the area of the fire grate, calculating the circularity of the edge of the fire grate according to the area of the fire grate and the length of the edge of the fire grate, and using the circularity as a shape characteristic.
The circularity of the edge profile of the boiler grate is analyzed to obtain the shape index O of the boiler grate:
s represents the area of the fire grate and is obtained by counting the number of pixel points in the edge of the fire grate; and p represents the perimeter of the edge of the fire grate and is obtained by counting the number of pixel points on the edge of the fire grate.
And S002, calculating the confidence coefficient of each pixel point gradient in the surface image, and performing Hough line detection on the pixel points with the confidence coefficient higher than a preset threshold value to obtain a plurality of coordinate points mapped in Hough space by a plurality of lines.
The method comprises the following specific steps:
1. and calculating the gradient amplitude of each pixel by utilizing the gradient of each pixel point in the horizontal direction and the vertical direction, and calculating the confidence coefficient according to the gradient amplitude.
Calculating the gradient amplitude G of each pixel point,
G x representing the gradient of the pixel in the horizontal direction, G y The gradient of the pixel points in the vertical direction is represented, and the fact that straight lines appear at positions with larger gradients is considered, so that the pixel points which are possibly straight lines are screened out by calculating the confidence coefficient w of the pixel points, and the confidence coefficient of the pixel points is calculatedThe method comprises the following steps:
based on the confidence coefficient of the pixel point, the pixel point in the surface image is filtered to reduce the detection amount in the Hough transform process and improve the transform speed, and the specific method comprises the following steps: setting a preset threshold value w T When the confidence of the pixel point is not higher than the preset threshold value w T And meanwhile, the confidence coefficient of the pixel point is low, and the conversion of the parameter space is not carried out on the pixel point when Hough transformation is carried out, so that the operation complexity is reduced.
As an example, in the embodiment of the present invention, w T The value is 0.4.
2. And carrying out Hough straight line detection on the pixel points with the reliability higher than a preset threshold value to obtain a plurality of coordinate points mapped in Hough space by a plurality of straight lines.
And only carrying out Hough transformation on pixel points with the reliability higher than a preset threshold value to realize straight line detection and obtain a plurality of straight lines and a plurality of coordinate points mapped in Hough space.
The confidence screening pixel points can be used for quickly performing preliminary evaluation on each pixel point in the grate surface image so as to obtain pixel points with higher confidence for Hough transformation and accurately extracting straight lines belonging to the grate.
S003, selecting the straight line segment with the longest length as a first grate bar, and acquiring a corresponding first coordinate point; screening out coordinate points with the same abscissa as the first coordinate points to serve as suspected grate points, and taking the mode of the distance between adjacent suspected grate points as a grate interval; and searching according to the grate intervals by taking the first coordinate point as a center to obtain all grate points.
The method comprises the following specific steps:
1. and selecting a first grate bar.
In the Hough space, a vertical distance d from an original point to a straight line segment is taken as a vertical coordinate, an included angle theta between a vertical line from the original point to the straight line segment and the horizontal line is taken as an abscissa, and a coordinate point (d, theta, l) corresponding to each straight line segment is obtained, wherein (d, theta) represents a coordinate, l represents a numerical value of the coordinate point, and the numerical value represents the length of the corresponding straight line segment. Obtaining a numerical value sequence { l1, l2, …, lv } of all points in the Hough space, wherein v is the number of coordinate points in the Hough space, namely the number of detected straight line segments in the grate surface image, and lv is the numerical value of a vth point.
Selecting the coordinate point with the maximum value in the numerical value sequence as a first coordinate point, and recording as (d) m ,θ m ,l m ) The corresponding straight line segment is used as a first grate bar. Since the numerical value represents the length of the straight line segment, the straight line segment with the longest length must be the grate bar belonging to the grate, and therefore the straight line segment determined as the grate bar in the image can be found.
2. And acquiring the grate interval.
As the fire grate bars of the fire grate are parallel, in the Hough space, the abscissa of the coordinate point corresponding to all the fire grate bars is the same abscissa, and the same abscissa theta with the first coordinate point is found m And the other coordinate points of (2) are taken as suspected fire points.
Because redundant attachments such as burrs and the like may exist on the grate bars, the attachments are even on the same abscissa theta m But not the fire grate points, the fire grate points are therefore captured by determining the grate spacing.
Acquiring a distance sequence { D1, D2, …, D (k-1) } between adjacent suspected fire grate points, wherein k represents the number of suspected fire grate points. Selecting the mode of the distance between adjacent suspected grate points as the grate interval dp:
dp＝Argmax(f Di )
wherein f is Di Indicating the frequency of occurrence of the ith distance Di.
3. And searching according to the grate intervals by taking the first coordinate point as a center to obtain all grate points.
At θ = θ m And searching the two sides of the straight line by taking the first coordinate point as the center and the grate interval as the searching distance, and obtaining coordinate points corresponding to the grate bars in pairs to obtain the grate points.
And counting the number u of the fire grate points including the first coordinate point, namely the real number of the fire grate bars in the surface image, comparing the number u with the number n of the fire grate bars of the standard fire grate, and judging whether redundant attachments exist on the fire grate or the fire grate bars are lost.
When u is less than n, the fire grate is considered to have the phenomenon of grate bar missing, so that the subsequent use is influenced; when u is larger than or equal to n, the fire grate is considered to have redundant attachments such as burrs and the like, or the fire grate bars are absent and the redundant attachments such as the burrs and the like are simultaneously generated.
It should be noted that, when coordinate points corresponding to the grate bars are obtained in pairs, when values l of corresponding points on two sides are different, the grate bar corresponding to the coordinate point with the smaller value may be damaged or lost.
Therefore, the grate bars in the surface image of the grate are detected, and the pixel points and the position information of the grate bars in the surface image are obtained, so that the defect area can be accurately extracted subsequently, and the system detection precision is improved.
And step S004, obtaining a semantic segmentation image of the surface image, extracting grate bars corresponding to the grate points from the semantic segmentation image, and performing reverse color on the grate bars and the edge of the grate to obtain a defect area.
Specifically, for the surface image, the fire grate pixels are detected and sensed through a semantic sensing network, the fire grate pixel points are marked as 1, the background pixel points are marked as 0, and the semantic segmentation image of the boiler fire grate can be obtained. And then, acquiring grate bar information in the surface image through the steps, setting the pixel values of the pixel points corresponding to the grate bars to zero, setting the pixel values of the pixel points at the edge of the grate to zero, and only reserving a defect area in the image to realize accurate extraction of the defect area.
The defect region is a region of an unnecessary deposit such as a burr, and a defect in which a grate bar is missing is already detected in step S003.
The method also includes the steps of:
1. local features of the defect region are obtained.
Firstly, a connected domain of a defect region is obtained, and the area, the position information and the distribution index of the defect region are used as local characteristics of the fire grate.
1.1 the area of the defect region is defined as the sum of the areas of the connected domains of all the defect regions.
1.2 acquiring the position information of the defect area.
Obtaining a center point position information sequence of a connected domain of z defect regions through a connected domain analysis algorithm { (x) 1 ,y 1 ),(x 2 ,y 2 ),…,(x z ,y z ) And constructing a defect position index analysis model, calculating a boiler grate defect position characteristic index, and analyzing the influence of the defect position on the abnormal detection of the boiler grate. The defect position index analysis model is as follows:
wherein (x) j ,y j ) Coordinates of center point of connected component representing jth defect area, (x) 0 ,y 0 ) Representing the coordinates of the center point of the boiler grate.
1.3, acquiring a distribution index of the defect area.
As an example, a distribution index obtaining method according to an embodiment of the present invention includes:
clustering the central points of all the defect areas to obtain the number of clustering categories and the number of discrete points; and calculating the distribution index according to the number of the cluster categories and the number of the discrete points.
The embodiment of the invention clusters the central point of the connected domain of the defect area by adopting a DBSCAN clustering algorithm, then statistically analyzes each cluster after clustering, and counts the number of the clustering centers as C 1 The number of discrete points is C 2 And calculating a distribution index Gat:
Gat＝w 1 *C 1 +w 2 *C 2
wherein, w 1 Weight, w, representing the number of cluster centers 2 A weight representing the number of discrete points.
In the examples of the present invention w 1 The value is 0.4,w 2 Is 0.6.
As another example, a distribution index obtaining method according to another embodiment of the present invention includes:
and acquiring the shortest distance between every two defect areas, and taking the sum of all the shortest distances as a distribution index.
The shortest distance between any two defect regions is obtained by calculating the distance between any two pixel points belonging to the two different defect regions, and then the shortest distance sequence d between every two defect regions is obtained min ＝{d m1 ,d m2 ,…,d mt }，
And representing the sequence length of the shortest distance sequence, and taking the sum of all shortest distances as a distribution index.
2. And calculating the abnormal degree of the fire grate according to the overall characteristics and the local characteristics.
And establishing a boiler grate abnormity detection model, and evaluating the abnormity degree of the boiler grate. The function model of the abnormal degree of the fire grate is specifically as follows:
wherein, U n Indicates the degree of abnormality, S L Indicating the area of the defective region.
U n The larger the value of the value is, the higher the abnormal degree of the fire grate is, normalization processing is carried out on the abnormal degree, and the value range of the function value is guaranteed to be (0,1), so that the abnormal degree of the boiler fire grate can be visually analyzed.
When the function value of the function model of the degree of abnormality is higher than 0.3, it is determined that the boiler grate is in an abnormal condition, and an operator needs to perform processing again.
The embodiment of the invention adopts the conveyor belt device to realize the full automation of the detection of the boiler grate, the manipulator is arranged at one side of the tail part of the conveyor belt, when the method of the invention detects that the boiler grate is abnormal, the detection result is sent to the manipulator control device, the manipulator is started to place the abnormal boiler grate to the abnormal part storage position, and the abnormal boiler grate is separated from the boiler grate which can be put into use, so that the related operators can process the abnormal grate again.
In summary, in the embodiment of the present invention, the surface image of the grate is collected at the overlooking view angle, and the grate edge of the grate is extracted; calculating the confidence coefficient of the gradient of each pixel point in the surface image, and performing Hough line detection on the pixel points with the confidence coefficient higher than a preset threshold value to obtain a plurality of coordinate points mapped in Hough space by a plurality of lines; selecting a straight line segment with the longest length as a first grate bar, and acquiring a corresponding first coordinate point; screening out coordinate points with the same abscissa as the first coordinate points to serve as suspected grate points, and taking the mode of the distance between adjacent suspected grate points as a grate interval; searching according to the grate intervals by taking the first coordinate point as a center to obtain all grate points; and obtaining a semantic segmentation image of the surface image, extracting grate bars corresponding to the grate points from the semantic segmentation image, and performing reverse color on the grate bars and the edge of the grate to obtain a defect area. The embodiment of the invention can accurately and quickly extract the defect area in the grate.
The embodiment of the invention also provides an artificial intelligence based fire grate abnormality detection system, which comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the processor realizes the steps of the method when executing the computer program. Since the grate abnormality detection method based on artificial intelligence is described in detail above, it is not described again.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And that specific embodiments have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
All the embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. The grate anomaly detection method based on artificial intelligence is characterized by comprising the following steps of:
collecting a surface image of a fire grate at a top view angle, and extracting the edge of the fire grate;
calculating the confidence coefficient of each pixel point gradient in the surface image, and performing Hough straight line detection on pixel points with the confidence coefficient higher than a preset threshold value to obtain a plurality of coordinate points mapped in Hough space by a plurality of straight lines;
selecting a straight line segment with the longest length as a first grate bar, and acquiring a corresponding first coordinate point; screening out coordinate points with the same abscissa as the first coordinate points to serve as suspected grate points, and taking the mode of the distance between adjacent suspected grate points as a grate interval; searching according to the grate intervals by taking the first coordinate point as a center to obtain all grate points;
and obtaining a semantic segmentation image of the surface image, extracting a grate bar corresponding to the grate point from the semantic segmentation image, and performing reverse color on the grate bar and the grate edge to obtain a defect area.
2. The artificial intelligence based fire grate abnormality detecting method according to claim 1, further comprising the steps of:
acquiring the integral characteristics of the fire grate according to the edge of the fire grate; the integral features include dimensional and shape features;
acquiring a distribution index of the defect area; taking the area and the position information of the defect area and the distribution index as local characteristics of the fire grate;
and calculating the abnormal degree of the fire grate according to the overall characteristic and the local characteristic.
3. The grate abnormality detection method based on artificial intelligence of claim 2, wherein the distribution index acquisition method is:
clustering the central points of all the defect areas to obtain the number of clustering categories and the number of discrete points;
and calculating the distribution index according to the number of the cluster categories and the number of the discrete points.
4. The grate abnormality detection method based on the artificial intelligence, according to claim 2, wherein the distribution index acquisition method is:
and acquiring the shortest distance between every two defect areas, and taking the sum of all the shortest distances as the distribution index.
5. The artificial intelligence based grate anomaly detection method of claim 2, wherein the step of obtaining dimensional characteristics includes:
and obtaining the detection radius of the fire grate according to the edge of the fire grate, and taking the difference value of the detection radius and the standard radius of a standard fire grate as the size characteristic.
6. The artificial intelligence based grate abnormality detection method according to claim 2, wherein the shape feature obtaining step includes:
and counting the number of pixel points in the edge of the fire grate to be used as the area of the fire grate, calculating the circularity of the edge of the fire grate according to the area of the fire grate and the length of the edge of the fire grate, and using the circularity as the shape characteristic.
7. The method for detecting grate abnormality based on artificial intelligence as claimed in claim 1, wherein the confidence coefficient is calculated by:
and calculating the gradient amplitude of each pixel by utilizing the gradient of each pixel point in the horizontal direction and the vertical direction, and calculating the confidence coefficient according to the gradient amplitude.
8. The method for detecting abnormal fire grates based on artificial intelligence of claim 5, wherein the step of obtaining all fire grate points further comprises the steps of:
and counting the number of the fire grate points, comparing the number with the number of the fire grate bars of the standard fire grate, and judging whether the fire grate has redundant attachments or lacks the fire grate bars.
9. The method for detecting the abnormal fire grate based on the artificial intelligence as claimed in claim 1, wherein the step of extracting the edge of the fire grate further comprises:
and removing the noise of the surface image to obtain a de-noised image, and then carrying out illumination equalization processing on the de-noised image.
10. The system for detecting the abnormal fire grate based on the artificial intelligence comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the processor executes the computer program to realize the steps of the method for detecting the abnormal fire grate based on the artificial intelligence according to any one of the claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210346787.1A CN114782330B (en) | 2022-03-31 | 2022-03-31 | Grate abnormity detection method and system based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210346787.1A CN114782330B (en) | 2022-03-31 | 2022-03-31 | Grate abnormity detection method and system based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114782330A CN114782330A (en) | 2022-07-22 |
| CN114782330B true CN114782330B (en) | 2022-12-16 |
Family
ID=82427770
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210346787.1A Active CN114782330B (en) | 2022-03-31 | 2022-03-31 | Grate abnormity detection method and system based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114782330B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115526889B (en) * | 2022-11-24 | 2023-03-03 | 江苏太湖锅炉股份有限公司 | Nondestructive testing method of boiler pressure pipeline based on image processing |
| CN116642893B (en) * | 2023-07-24 | 2023-10-03 | 吉林省艾优数字科技有限公司 | Visual intelligent detection method, device, equipment and medium for antigen detection reagent |
| CN117078661B (en) * | 2023-10-12 | 2024-01-05 | 济宁久邦工程机械设备有限公司 | Intelligent detection method for production quality of structural member of overhead working truck |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10157081A (en) * | 1996-11-29 | 1998-06-16 | Dainippon Printing Co Ltd | Method and equipment for inspecting streaky defect |
| JP2006153614A (en) * | 2004-11-29 | 2006-06-15 | Nippon Steel Corp | Method and apparatus for detecting cyclic flaw of strip-like member or columnar member |
| JP2007057705A (en) * | 2005-08-23 | 2007-03-08 | Seiko Epson Corp | Inspection method and inspection apparatus of display panel |
| JP2010243263A (en) * | 2009-04-02 | 2010-10-28 | Nippon Steel Corp | Surface inspecting device and surface inspection method of belt-like body, and program |
| JP2011118509A (en) * | 2009-12-01 | 2011-06-16 | Fuji Heavy Ind Ltd | Road recognition device |
| WO2021068497A1 (en) * | 2019-10-10 | 2021-04-15 | 南京凯盛国际工程有限公司 | Grate cooler working state monitoring method based on image recognition |
| CN114078107A (en) * | 2020-08-11 | 2022-02-22 | 华晨宝马汽车有限公司 | Method, storage medium, device and system for identifying strip-shaped trim of vehicle |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR102227843B1 (en) * | 2014-08-22 | 2021-03-15 | 현대모비스 주식회사 | Operating method of lane departure warning system |
| JP7247515B2 (en) * | 2018-10-23 | 2023-03-29 | コニカミノルタ株式会社 | Image inspection device and image inspection program |
-
2022
- 2022-03-31 CN CN202210346787.1A patent/CN114782330B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH10157081A (en) * | 1996-11-29 | 1998-06-16 | Dainippon Printing Co Ltd | Method and equipment for inspecting streaky defect |
| JP2006153614A (en) * | 2004-11-29 | 2006-06-15 | Nippon Steel Corp | Method and apparatus for detecting cyclic flaw of strip-like member or columnar member |
| JP2007057705A (en) * | 2005-08-23 | 2007-03-08 | Seiko Epson Corp | Inspection method and inspection apparatus of display panel |
| JP2010243263A (en) * | 2009-04-02 | 2010-10-28 | Nippon Steel Corp | Surface inspecting device and surface inspection method of belt-like body, and program |
| JP2011118509A (en) * | 2009-12-01 | 2011-06-16 | Fuji Heavy Ind Ltd | Road recognition device |
| WO2021068497A1 (en) * | 2019-10-10 | 2021-04-15 | 南京凯盛国际工程有限公司 | Grate cooler working state monitoring method based on image recognition |
| CN114078107A (en) * | 2020-08-11 | 2022-02-22 | 华晨宝马汽车有限公司 | Method, storage medium, device and system for identifying strip-shaped trim of vehicle |
Non-Patent Citations (5)
| Title |
|---|
| A Machine Vision System for Defect Detection of a Traveling Grate Conveyor;Ahmad Pouranimi 等;《2015 2nd International Conference on Knowledge-Based Engineering and Innovation》;20160321;1063-1066 * |
| Combining Hough transform and contour algorithm for detecting vehicles" license-plates;Tran Due Duan 等;《Proceedings of 2004 International Symposium on Intelligent Multimedia, Video and Speech Processing》;20050606;747-750 * |
| Image Processing Based Defect Detection and Recognition of Embossed Characters in Induration Conveyor for Pellet Plant;Avi Bhanja 等;《International Journal of Recent Engineering Research and Development》;20210331;48-59 * |
| 圆形声纳浮标阵应召搜潜效能仿真研究;樊俊，谢国新;《System Simulation Technology & Application》;20120811;400-403 * |
| 触点材料表面缺陷检测技术的研究;杜丹阳;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20210115;C042-1215 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114782330A (en) | 2022-07-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114782330B (en) | Grate abnormity detection method and system based on artificial intelligence | |
| CN110163853B (en) | Edge defect detection method | |
| CN115082418B (en) | Precise identification method for automobile parts | |
| CN116205919B (en) | Hardware part production quality detection method and system based on artificial intelligence | |
| CN115330645B (en) | Welding image enhancement method | |
| CN107610104A (en) | Crack detecting method at a kind of elevator compensation chain R based on machine vision | |
| CN109636824A (en) | A kind of multiple target method of counting based on image recognition technology | |
| CN110866903A (en) | Ping-pong ball identification method based on Hough circle transformation technology | |
| CN116664559B (en) | Machine vision-based memory bank damage rapid detection method | |
| CN113870235A (en) | Method for detecting defects of circular stamping part based on quantum firework arc edge extraction | |
| CN115797354B (en) | Method for detecting appearance defects of laser welding seam | |
| CN109781737B (en) | Detection method and detection system for surface defects of hose | |
| CN111476804B (en) | Efficient carrier roller image segmentation method, device, equipment and storage medium | |
| CN110415208A (en) | A kind of adaptive targets detection method and its device, equipment, storage medium | |
| CN116109637B (en) | System and method for detecting appearance defects of turbocharger impeller based on vision | |
| CN112541889A (en) | Method for detecting surface defects of complex texture object of multi-level model | |
| CN115063430A (en) | Electric pipeline crack detection method based on image processing | |
| CN113780110A (en) | Method and device for detecting weak and small targets in image sequence in real time | |
| CN107516315B (en) | Tunneling machine slag tapping monitoring method based on machine vision | |
| CN114119603A (en) | Image processing-based snack box short shot defect detection method | |
| CN116524205A (en) | Sewage aeration automatic detection and identification method | |
| CN115171218A (en) | Material sample feeding abnormal behavior recognition system based on image recognition technology | |
| CN114663427A (en) | Boiler part size detection method based on image processing | |
| CN115018785A (en) | Hoisting steel wire rope tension detection method based on visual vibration frequency identification | |
| CN112288760B (en) | Adherent cell image screening method and system and cell image analysis method |
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 |