CN114235814A - Crack identification method for building glass curtain wall - Google Patents
Crack identification method for building glass curtain wall Download PDFInfo
- Publication number
- CN114235814A CN114235814A CN202111461292.5A CN202111461292A CN114235814A CN 114235814 A CN114235814 A CN 114235814A CN 202111461292 A CN202111461292 A CN 202111461292A CN 114235814 A CN114235814 A CN 114235814A
- Authority
- CN
- China
- Prior art keywords
- image
- curtain wall
- glass curtain
- cracks
- detected
- 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.)
- Granted
Links
- 239000011521 glass Substances 0.000 title claims abstract description 119
- 238000000034 method Methods 0.000 title claims abstract description 33
- 239000013598 vector Substances 0.000 claims abstract description 18
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000013145 classification model Methods 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims abstract description 5
- 238000001514 detection method Methods 0.000 claims description 16
- 238000003708 edge detection Methods 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 239000005328 architectural glass Substances 0.000 claims 6
- 230000007547 defect Effects 0.000 description 8
- 239000005341 toughened glass Substances 0.000 description 8
- 238000000605 extraction Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000012706 support-vector machine Methods 0.000 description 5
- 238000013459 approach Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 208000037656 Respiratory Sounds Diseases 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 241000239290 Araneae Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 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
- 230000001681 protective effect Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- E—FIXED CONSTRUCTIONS
- E04—BUILDING
- E04B—GENERAL BUILDING CONSTRUCTIONS; WALLS, e.g. PARTITIONS; ROOFS; FLOORS; CEILINGS; INSULATION OR OTHER PROTECTION OF BUILDINGS
- E04B2/00—Walls, e.g. partitions, for buildings; Wall construction with regard to insulation; Connections specially adapted to walls
- E04B2/88—Curtain walls
-
- E—FIXED CONSTRUCTIONS
- E04—BUILDING
- E04B—GENERAL BUILDING CONSTRUCTIONS; WALLS, e.g. PARTITIONS; ROOFS; FLOORS; CEILINGS; INSULATION OR OTHER PROTECTION OF BUILDINGS
- E04B2/00—Walls, e.g. partitions, for buildings; Wall construction with regard to insulation; Connections specially adapted to walls
- E04B2/88—Curtain walls
- E04B2/885—Curtain walls comprising a supporting structure for flush mounted glazing panels
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
Landscapes
- Engineering & Computer Science (AREA)
- Architecture (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method for identifying cracks of a building glass curtain wall, which comprises the following steps: collecting a glass curtain wall image; preprocessing the glass curtain wall image to obtain an image to be detected; extracting edges in the image to be detected to obtain an edge image; extracting a feature vector of the edge image, wherein the feature vector comprises a distribution density feature and an on pixel feature; constructing and training a classification model; and inputting the characteristic vector of the edge image into a classification model to obtain a classification result, wherein the classification result comprises cracks and non-cracks.
Description
Technical Field
The invention relates to a crack identification method for a building glass curtain wall, and belongs to the field of glass crack identification.
Background
The glass curtain wall is a building external protective structure or a decorative structure which has a certain displacement capacity relative to the main structure by a supporting structure system and does not bear the action of the main structure. Most of glass curtain walls are made of toughened glass materials, but the toughened glass is brittle, is easy to crack and even falls off from high altitude, endangers the life safety of pedestrians, and needs to be maintained and repaired regularly. The maintenance and overhaul of the existing glass curtain wall need a plurality of professional workers with high-altitude operation qualification to carry out manual overhaul through equipment such as an electric hanging basket, and the maintenance and overhaul method is low in efficiency and high in cost. There is a need for a method of automatically identifying and locating cracks in a glass curtain wall.
Patent publication No. CN109668909A, a glass defect detection method, discloses the following steps: (1) image acquisition: acquiring a glass image on a production line in real time by adopting a high-speed linear array CCD camera, and carrying out digital processing on an acquired image analog signal through an image acquisition card and then transmitting the image analog signal to a computer for image preprocessing; (2) image preprocessing: the method comprises the steps of image filtering processing and image enhancement processing; (3) image segmentation: distinguishing different regions having special meanings therein according to the gray scale, color or geometric properties of the image; (4) feature extraction: determining characteristic parameters and extracting glass defect characteristics; (5) and (3) judging and deciding: classifying the glass defects according to the step (4). The characteristic parameters of the selected glass defects are length-diameter ratio, perimeter-square area ratio or ratio of area pixel number to perimeter pixel number.
Patent publication CN102305798A, entitled "method for detecting and classifying glass defects based on machine vision", includes first extracting the defect region in the picture given by a camera (line scanning) by using Canny edge detection, so as to obtain the minimum connected domain of the defect. The target area is then processed according to the proposed filter and W-feature of the invention. And then defining 9 types of characteristic modes, scanning the minimum connected domain according to rows and columns, counting the frequency of the 9 types of characteristic modes in the sample, and judging the type of the defect on the basis.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides a method for identifying cracks of a building glass curtain wall, which designs a characteristic vector with the distribution density characteristic and the on-pixel characteristic as edges, can effectively extract the cracks in the glass curtain wall and reduce misjudgment caused by stain interference.
The technical scheme of the invention is as follows:
a method for identifying cracks of a building glass curtain wall comprises the following steps:
collecting a glass curtain wall image;
preprocessing the glass curtain wall image to obtain an image to be detected;
extracting edges in the image to be detected to obtain an edge image;
extracting a feature vector of the edge image, wherein the feature vector comprises a distribution density feature and an on-pixel feature;
constructing and training a classification model; and inputting the characteristic vector of the edge image into a classification model to obtain a classification result, wherein the classification result comprises cracks and non-cracks.
Further, the collecting of the glass curtain wall image specifically comprises: shooting the glass curtain wall through the unmanned aerial vehicle provided with the image acquisition device to obtain a glass curtain wall image.
Further, the method also comprises the following steps: presetting a flight route; the unmanned aerial vehicle shoots all glass curtain walls on the building in sequence according to the aircraft route and numbers the glass curtain walls to obtain a glass curtain wall image set.
Further, the method also comprises the following steps:
preprocessing the glass curtain wall image set to obtain a corresponding image set to be detected;
carrying out feature detection and feature matching on the image set to be detected; splicing the image sets to be detected according to the matched feature pairs to obtain a curtain wall panoramic image;
and determining the position of the glass curtain wall with the crack in the panoramic picture of the curtain wall according to the serial number of the glass curtain wall.
Further, the preprocessing of the glass curtain wall image specifically comprises:
obtaining the position of a glass frame in the image of the glass curtain wall through a Hough line detection algorithm; and performing oblique correction on the glass curtain wall image according to the frame position through a central projection transformation algorithm to obtain an image to be detected.
Further, the extracting the edge image specifically includes:
dividing the image to be detected into a plurality of areas by using a clustering algorithm; and respectively carrying out edge detection on the plurality of areas to obtain edge images.
Further, edge detection is performed by using a canny operator, and a first threshold value and a second threshold value, which are used for filtering edge points, of the canny operator are obtained by a maximum inter-class difference method, wherein the method specifically comprises the following steps:
traversing the image to be detected, solving a threshold value T which enables the variance value between the maximum classes to be maximum,
taking a threshold T as a first threshold, and taking half of the threshold T as a second threshold.
The invention has the following beneficial effects:
1. according to the invention, through observation and statistics of a large number of glass samples, cracks generated by toughened glass used for the glass curtain wall are more in a spider-web shape, the crack density and the area are far greater than those generated by other non-toughened glass, and the crack is obviously different from stains, has the characteristics of non-uniform gray level, large difference with background gray level, large proportion of target pixel points and the like, so that the characteristic vectors with the distribution density characteristics and the on pixel characteristics as edges are designed, cracks in the glass curtain wall can be effectively extracted, and misjudgment caused by stain interference is reduced.
2. According to the invention, all images to be detected of the building are spliced to obtain the curtain wall panoramic view, and the integral crack condition of the building glass curtain wall can be intuitively known by positioning the position of the glass curtain wall with cracks by utilizing the curtain wall number determined in the shooting process in advance.
3. The invention carries out clustering treatment on the image to be detected and then carries out edge detection, can reduce the condition of missing detection on the crack with gradually changed gray level at the glass curtain wall and improve the completeness and continuity of the detected edge.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a schematic view of a flight path of an unmanned aerial vehicle;
FIG. 3 is a schematic view of the numbering of the glass curtain wall;
FIG. 4 is a schematic view of an image to be inspected;
fig. 5 is a schematic diagram of an edge image.
Detailed Description
The invention is described in detail below with reference to the figures and the specific embodiments.
Example one
Referring to fig. 1, a method for identifying cracks of a building glass curtain wall comprises the following steps:
collecting a glass curtain wall image;
preprocessing the glass curtain wall image to obtain an image to be detected;
extracting edges in the image to be detected to obtain an edge image;
extracting a feature vector of the edge image, wherein the feature vector comprises a distribution density feature and an on-pixel feature;
constructing and training a classification model; and inputting the characteristic vector of the edge image into a classification model to obtain a classification result, wherein the classification result comprises cracks and non-cracks.
The beneficial effect of this embodiment lies in, through observation and statistics to a large amount of glass samples, the crack that the toughened glass that discovery glass curtain wall used produced is with spider web shape many, no matter crack density or area all is far greater than the crack that other non-toughened glass produced, and have showing difference with the stain, have that the grey scale is inhomogeneous, with background grey value difference big, target pixel takes up than characteristics such as big, so design distribution density characteristic and on pixel characteristic are the eigenvector of edge, can effectively extract the crackle in the glass curtain wall and reduce the erroneous judgement that the stain interference caused.
Example two
The flight path of the unmanned aerial vehicle is preset.
The unmanned aerial vehicle utilizes the image acquisition device that its carried to shoot all glass curtain walls on the building in proper order and serial number glass curtain wall according to the aircraft route, obtains glass curtain wall image collection.
And preprocessing the image set of the glass curtain wall to obtain a corresponding image set to be detected.
Carrying out crack identification on the image set to be detected to obtain a glass curtain wall with cracks;
carrying out feature detection and feature matching on the image set to be detected; splicing the image sets to be detected according to the matched feature pairs to obtain a curtain wall panoramic image;
and determining the position of the glass curtain wall with the crack in the panoramic picture of the curtain wall according to the serial number of the glass curtain wall.
The beneficial effect of this embodiment lies in, through all waiting to examine the image of concatenation building, acquires the curtain wall panorama to utilize the curtain serial number of confirming at the shooting in-process in advance, the whole crackle condition of building glass curtain wall is known directly perceivedly to the position that the location appears of band crack glass curtain wall.
EXAMPLE III
A method for identifying cracks of a building glass curtain wall comprises the following steps:
step 1: collecting glass curtain wall image
As shown in fig. 2 and 3, the unmanned aerial vehicle based on the GPS signal is adopted to inspect the outer facade of the glass curtain wall of the building. According to the characteristics of the outer facade of the building glass curtain wall, the curtain wall is divided into a plurality of areas, and the acquisition positions of images in the areas are determined. In order to ensure the image quality, the unmanned aerial vehicle is close to the outer vertical surface of the glass curtain wall as much as possible, and a certain overlapping part is required between images which contain a piece of complete glass and adjacent glass curtain walls in a single glass curtain wall image. In addition, considering that the transparency and the light reflection of the glass can influence the crack detection accuracy, the unmanned aerial vehicle carries a camera to collect the images of the glass curtain wall in a mode of upward shooting, and numbers the glass curtain wall according to the sequence of the shot images.
Step 2: preprocessing glass curtain wall images
When the unmanned aerial vehicle collects the photos, the acquired images have certain deformation under the influence of various factors such as wind speed, illumination, height and the like. In the embodiment, Hough line detection is carried out on the glass curtain wall image to extract coordinates of corner points of the glass frame, so that the glass curtain wall image is subjected to inclination correction.
Step 2.1, extracting a single piece of glass based on Hough line detection
Glass curtain walls are usually composed of glass and a glass frame supporting the glass, which can be seen as a rectangle enveloping the glass, with a significant difference compared to the glass itself. The position of the frame can be determined through Hough line detection, and the edge of the glass can be indirectly determined, so that the extraction of a single piece of glass is realized.
The Hough line detection algorithm has strong anti-interference capability, is insensitive to the incomplete part of a straight line in an image, noise and other coexisting nonlinear structures, can tolerate a gap in feature boundary description, and is relatively free from the influence of image noise, and comprises the following specific steps of:
1) and carrying out binarization on the glass curtain wall image. And carrying out Hough line detection on the binarized glass curtain wall image, and storing the detected edge point coordinates.
2) And substituting the detected edge point coordinates into a formula (1), and solving Hough line detection parameters (m, a).
Wherein (x)1,y1),(x2,y2) As arbitrary in the imageTwo edge points.
3) If the formula (1) has a solution, it indicates that a straight line exists between the two edge points (the value m is the slope of the straight line, and the value a is the intercept of the straight line). And counting the number of all the straight lines, and deleting the straight lines of which the number is smaller than a preset threshold value. The remaining straight line is a straight line indicating the position of the glass frame.
4) And extracting the intersection points of the straight lines to obtain the coordinates of the four corner points of the glass frame.
Step 2.2: tilt correction
According to the invention, the image is acquired by adopting a mode of overhead shooting of the unmanned aerial vehicle, the acquired image has deformation to a certain extent, and in order to avoid influence of perspective distortion of the image on subsequent crack extraction, after the image is acquired by the unmanned aerial vehicle, the four angular point coordinates obtained in the step 2.1 are utilized, and perspective transformation is adopted to perform inclination correction on the image of the glass curtain wall, so that the image to be detected is obtained, as shown in fig. 4.
And step 3: crack detection
Step 3.1: it is considered that there are disturbance factors such as stains on the glass in addition to the cracks. Therefore, the K-means algorithm is used to divide the image to be inspected into three areas (a crack area, a stain area and a healthy area), and the method comprises the following steps:
1) and respectively selecting 3 pixel points as initial clustering centers.
2) And calculating the distance from each pixel in the image to be detected to the 3 cluster centers, and classifying the pixel into the cluster center with the minimum distance.
3) The cluster center for each class is recalculated.
4) And (5) repeating the steps 2 and 3 until the position of the cluster center is not changed. Finally obtaining a crack area, a stain area and a healthy area.
Step 3.2: respectively carrying out edge detection on the crack area, the stain area and the healthy area by utilizing a canny operator, and specifically comprising the following steps:
1) and smoothing each region by using a Gaussian filter respectively to filter noise.
2) And respectively calculating the gradient strength and the direction of each pixel point in each region to obtain a corresponding gradient image.
3) And respectively carrying out non-maximum suppression on all pixels in each gradient image so as to eliminate stray response brought by edge detection.
4) Calculating the highest threshold and the lowest threshold by a maximum inter-class variance method: and traversing the image to be detected, solving a threshold T which enables the variance value between the maximum classes to be maximum, taking the threshold T as a first threshold, and taking half of the threshold T as a second threshold. And further screening the edge points according to the highest threshold and the lowest threshold.
5) Isolated weak edges are suppressed.
Step 3.3: the crack area, stain area, and healthy area are stitched into an edge image, as shown in fig. 5.
Step 3.4: support Vector Machine (SVM) -based true and false crack identification and classification
In order to filter the smudge edges influencing crack extraction, crack identification is carried out through an SVM classifier, and glass crack identification is carried out according to feature classification and machine learning. In the crack classification process, the feature extraction is digital information for constructing and distinguishing cracks and stains of the glass curtain wall, and the method is mainly used for classifying and identifying through the following two types of features:
1) distribution density characteristics
Calculating the ratio R of the number of non-0-value pixels in each line of the edge image to the total number of line elements in a line uniti;
Preset threshold K1 and threshold K2 (in the present embodiment, K1 is 0.1, and K2 is 0.4);
comparing R corresponding to each rowiThe values and thresholds K1, K2; statistics of RiNumber of rows, < K1, noted as L1(ii) a Statistics K1 < RiNumber of rows, < K2, noted as L2(ii) a Statistics K2 < RiThe number of lines of (1) is noted as L3。
To L1、L2、L3Performing normalization processing to calculate L1、L2、L3The ratio of the number of the total lines is obtained as distribution density feature vectors feature1, feature2 and feature3 corresponding to the number of elements of 3.
Since the crack distribution of the tempered glass curtain wall almost occupies the whole glass, the feature1 and feature2 in the case of breakage of the tempered glass are expected to approach 0, and the value of feature3 approaches 1. For the case of glass without cracks and stains, feature1 is expected to approach 1, and features 2, 3 approach 0. For the case of glass that is intact but smudged, feature1, feature2, feature3 may all have values due to uncertainty in the smudge distribution, but are certainly distinctly different from the characteristics of cracked and non-cracked glass. And training a support vector machine by using sample data, and finding an optimal hyperplane on a feature space to enable the interval between the features of the cracked glass and the features of the non-cracked glass to be maximum, thereby accurately distinguishing the cracked glass from the non-cracked glass.
2) on pixel feature
And taking the pixels with the pixel values not being zero in the binarized edge image as on pixels. The total area of cracks and stains in the binarized edge image, i.e. the number of on pixels in the image, was estimated by the bwearea function in Matlab. The ratio feature4 of the number of on pixels to the total number of picture pixels is calculated as a feature vector. The test shows that the state of the glass curtain wall can be effectively reflected.
And extracting feature vectors (feature1, feature2, feature3 and feature4) of the edge images, inputting the feature vectors into the trained SVM model, and obtaining a classification result (crack or non-crack), so that whether the crack exists in the corresponding glass curtain wall can be known.
And 4, step 4: crack location
The building curtain wall panoramic image is formed by splicing a plurality of images with overlapped areas at different angles. After crack identification is carried out on the image, in order to accurately determine the specific position of the cracked glass curtain wall, feature points of the image collected by the unmanned aerial vehicle are described by adopting an SIFT algorithm, effective matching feature points are screened by using a FLANN method, and invalid matching feature points are filtered. And splicing the images to be detected according to the matching of the effective characteristic points to obtain a curtain wall panoramic image. The position of the glass curtain wall with cracks can be quickly determined according to the preset number information of the glass curtain wall.
The method has the advantages that clustering processing is carried out on the image to be detected, then edge detection is carried out, the condition of missing detection of the crack with gradually changed gray level at the glass curtain wall can be reduced, and the integrity and the continuity of the detected edge are improved.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or applied directly or indirectly to other related technical fields, are included in the scope of the present invention.
Claims (7)
1. A method for identifying cracks of a building glass curtain wall is characterized by comprising the following steps:
collecting a glass curtain wall image;
preprocessing the glass curtain wall image to obtain an image to be detected;
extracting edges in the image to be detected to obtain an edge image;
extracting a feature vector of the edge image, wherein the feature vector comprises a distribution density feature and an on pixel feature;
constructing and training a classification model; and inputting the characteristic vector of the edge image into a classification model to obtain a classification result, wherein the classification result comprises cracks and non-cracks.
2. The method for identifying the cracks of the architectural glass curtain wall according to claim 1, wherein the collecting of the glass curtain wall images specifically comprises: shooting the glass curtain wall through the unmanned aerial vehicle provided with the image acquisition device to obtain a glass curtain wall image.
3. The method for identifying cracks on an architectural glass curtain wall as claimed in claim 2, further comprising: presetting a flight route; the unmanned aerial vehicle shoots all glass curtain walls on the building in sequence according to the aircraft route and numbers the glass curtain walls to obtain a glass curtain wall image set.
4. The method for identifying cracks on an architectural glass curtain wall as claimed in claim 3, further comprising:
preprocessing the glass curtain wall image set to obtain a corresponding image set to be detected;
carrying out feature detection and feature matching on the image set to be detected; splicing the image sets to be detected according to the matched feature pairs to obtain a curtain wall panoramic image;
and determining the position of the glass curtain wall with the crack in the panoramic picture of the curtain wall according to the serial number of the glass curtain wall.
5. The method for identifying the cracks of the architectural glass curtain wall according to claim 1, wherein the preprocessing of the glass curtain wall image specifically comprises:
obtaining the position of a glass frame in the image of the glass curtain wall through a Hough line detection algorithm; and performing oblique correction on the glass curtain wall image according to the frame position through a central projection transformation algorithm to obtain an image to be detected.
6. The method for identifying the cracks of the architectural glass curtain wall as claimed in claim 1, wherein the extracting the edge image specifically comprises:
dividing the image to be detected into a plurality of areas by using a clustering algorithm; and respectively carrying out edge detection on the plurality of areas to obtain edge images.
7. The method for identifying cracks on an architectural glass curtain wall according to claim 6, wherein edge detection is performed by using a canny operator, and a first threshold value and a second threshold value used for filtering edge points by the canny operator are obtained by a maximum inter-class difference method, and specifically:
traversing the image to be detected, solving a threshold value T which enables the variance value between the maximum classes to be maximum,
taking a threshold T as a first threshold, and taking half of the threshold T as a second threshold.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111461292.5A CN114235814B (en) | 2021-12-02 | Crack identification method for building glass curtain wall |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111461292.5A CN114235814B (en) | 2021-12-02 | Crack identification method for building glass curtain wall |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114235814A true CN114235814A (en) | 2022-03-25 |
CN114235814B CN114235814B (en) | 2024-07-16 |
Family
ID=
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114881961A (en) * | 2022-04-29 | 2022-08-09 | 江苏仙岳材料科技有限公司 | Glass fiber plate crack detection method and device |
CN115660647A (en) * | 2022-11-05 | 2023-01-31 | 一鸣建设集团有限公司 | Maintenance method for building outer wall |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011058939A (en) * | 2009-09-09 | 2011-03-24 | Panasonic Electric Works Co Ltd | Apparatus and method for visual inspection |
CN101996405A (en) * | 2010-08-30 | 2011-03-30 | 中国科学院计算技术研究所 | Method and device for rapidly detecting and classifying defects of glass image |
CN102162797A (en) * | 2010-11-24 | 2011-08-24 | 哈尔滨工业大学(威海) | Algorithm for detecting glass bottle neck damage and bottle bottom dirt |
CN103473785A (en) * | 2013-09-29 | 2013-12-25 | 哈尔滨工业大学 | Rapid multiple target segmentation method based on three-valued image clustering |
CN111192194A (en) * | 2019-12-05 | 2020-05-22 | 同济大学 | Panoramic image splicing method for curtain wall building vertical face |
CN111929329A (en) * | 2020-07-30 | 2020-11-13 | 广东寰球智能科技有限公司 | Intelligent detection method and system for glass curtain wall and storage medium |
CN112927223A (en) * | 2021-03-29 | 2021-06-08 | 南通大学 | Glass curtain wall detection method based on infrared thermal imager |
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011058939A (en) * | 2009-09-09 | 2011-03-24 | Panasonic Electric Works Co Ltd | Apparatus and method for visual inspection |
CN101996405A (en) * | 2010-08-30 | 2011-03-30 | 中国科学院计算技术研究所 | Method and device for rapidly detecting and classifying defects of glass image |
CN102162797A (en) * | 2010-11-24 | 2011-08-24 | 哈尔滨工业大学(威海) | Algorithm for detecting glass bottle neck damage and bottle bottom dirt |
CN103473785A (en) * | 2013-09-29 | 2013-12-25 | 哈尔滨工业大学 | Rapid multiple target segmentation method based on three-valued image clustering |
CN111192194A (en) * | 2019-12-05 | 2020-05-22 | 同济大学 | Panoramic image splicing method for curtain wall building vertical face |
CN111929329A (en) * | 2020-07-30 | 2020-11-13 | 广东寰球智能科技有限公司 | Intelligent detection method and system for glass curtain wall and storage medium |
CN112927223A (en) * | 2021-03-29 | 2021-06-08 | 南通大学 | Glass curtain wall detection method based on infrared thermal imager |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114881961A (en) * | 2022-04-29 | 2022-08-09 | 江苏仙岳材料科技有限公司 | Glass fiber plate crack detection method and device |
CN115660647A (en) * | 2022-11-05 | 2023-01-31 | 一鸣建设集团有限公司 | Maintenance method for building outer wall |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2022042579A1 (en) | Lcd screen defect detection method and apparatus | |
CN110163853B (en) | Edge defect detection method | |
CN109141232B (en) | Online detection method for disc castings based on machine vision | |
US11774735B2 (en) | System and method for performing automated analysis of air samples | |
CN107255641B (en) | A method of Machine Vision Detection is carried out for self-focusing lens surface defect | |
CN105784713B (en) | Sealing ring detection method of surface flaw based on machine vision | |
CN110286124B (en) | Machine vision-based refractory brick measuring system | |
CN105388162B (en) | Raw material silicon chip surface scratch detection method based on machine vision | |
CN110033516B (en) | Needle flake particle content detection method based on binocular camera image acquisition and recognition | |
CN109900711A (en) | Workpiece, defect detection method based on machine vision | |
CN102426649B (en) | Simple steel seal digital automatic identification method with high accuracy rate | |
CN110261410A (en) | A kind of detection device and method of glass lens defect | |
CN107203990A (en) | A kind of labeling damage testing method based on template matches and image quality measure | |
CN111292321B (en) | Transmission line insulator defect image identification method | |
CN108364291A (en) | Grey cloth rapid detection method based on computer vision technique | |
Awrangjeb et al. | Improved building detection using texture information | |
JPH10506461A (en) | Method and apparatus for detecting cover slips on microscope slides | |
CN107610104A (en) | Crack detecting method at a kind of elevator compensation chain R based on machine vision | |
CN109559324A (en) | A kind of objective contour detection method in linear array images | |
WO2021000948A1 (en) | Counterweight weight detection method and system, and acquisition method and system, and crane | |
CN106780473A (en) | A kind of magnet ring defect multi-vision visual detection method and system | |
CN109345524A (en) | A kind of bearing open defect detection system of view-based access control model | |
CN111122590A (en) | Ceramic surface defect detection device and detection method | |
CN106529551B (en) | Intelligent recognition counting detection method for round-like objects in packaging industry | |
WO2023231262A1 (en) | Hoisting wire rope tension measurement method based on visual vibration frequency identification |
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 |