CN105389793A - Method for automatically identifying trend and width of fracture in object surface - Google Patents
Method for automatically identifying trend and width of fracture in object surface Download PDFInfo
- Publication number
- CN105389793A CN105389793A CN201510661260.8A CN201510661260A CN105389793A CN 105389793 A CN105389793 A CN 105389793A CN 201510661260 A CN201510661260 A CN 201510661260A CN 105389793 A CN105389793 A CN 105389793A
- Authority
- CN
- China
- Prior art keywords
- pixel
- curves
- point
- value
- crack
- 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
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
-
- 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
-
- 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
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8874—Taking dimensions of defect into account
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention relates to a method for automatically identifying the trend and the width of a fracture in an object surface. The method includes the following steps of: performing gray processing, binarization processing and filtering processing on an image, then performing Canny edge detection on the image to obtain an image of fracture edge pixel points; extracting two curves of a fracture edge and determining the trend of the fracture; and performing filtering processing on the two curves and calculating the width of the fracture. According to the method, the curves of the fracture edge in the object surface can be extracted by adoption of an image processing technology, the maximum width of the fracture can be calculated without manual adjustment of the trend of the fracture in the image, and the trend of the fracture can be automatically identified. The pixel points of the fracture edge can be reserved to the greatest extent through image processing by adoption of an image global binarization method based on a histogram, and the measurement accuracy can be improved; filtering processing is performed on the two curves of the fracture edge before calculation of the width of the fracture, the number of points in the curves is reduced, and the calculation speed of the width of the fracture can be improved.
Description
Technical field
The invention belongs to image processing field, be specifically related to a kind ofly utilize the automatic trend of recognition object surface crack of image processing techniques and the method for fracture width.
Background technology
The method of image processing techniques recognition object surface crack width is utilized mainly to comprise at present following several:
Artificial visually examine's method: adopt high precision camera to obtain crack pattern picture, utilize the rule on screen, differentiate the width in crack by the mode of range estimation.Artificial visually examine's method differentiates the width in crack, and differentiate that a crack needs tens seconds even a few minutes usually, and the subjectivity of personnel is large, the result that different people differentiates is all different.Measuring accuracy is low, cannot accomplish pixel scale.
Semi-automatic method of identification: the width utilizing computer image processing technology automatic discrimination crack.This method improves a lot in measuring accuracy relative to artificial visually examine's method, but semi-automatic method of identification need manually to adjust crack in the picture walk backward, just can automatically identify.Accuracy of measurement and crack trend in the picture also has relation.The inclination angle of fracture strike and vertical direction is less, and the accuracy of crack identification is higher.In order to ensure that accuracy of measurement needs adjustment crack trend in the picture repeatedly usually, this is the process of a very time-consuming.Therefore semi-automatic method of identification does not significantly improve in measurement efficiency relative to artificial visually examine's method.
Summary of the invention
In order to solve the above-mentioned problems in the prior art, the present invention proposes a kind of method of automatic recognition object surface crack trend and width, the method does not need the trend manually adjusting crack, as long as crack occurs in the picture, with regard to trend and the fracture width of the automatic crack identification of energy.
For achieving the above object, the present invention adopts following technical scheme:
A method for automatic recognition object surface crack trend and width, comprises the following steps:
Step 1, carries out gray processing, binaryzation and filtering process to image, then carries out Canny rim detection, obtain the image of edge of crack pixel.
Step 2, extracts two curves of edge of crack, and determines the trend in crack based on the path finding algorithm in the anti-game of tower.
Step 3, carries out filtering process to two curves of the edge of crack that step 2 obtains, and calculates fracture width.
Further, the image binaryzation process described in step 1 adopts a kind of based on histogrammic image overall binarization method, comprises the following steps:
(1) according to the gray-scale value compute histograms obtained after image gray processing process.
(2) if described histogram is a bimodal histogram, two peak-to-peak the lowest point values are got as threshold value; If described histogram is not a bimodal histogram, then to the smoothing process of histogram data.
(3) step (2) is repeated, until described histogram becomes a bimodal histogram.If do not obtain a bimodal histogram after being repeated N time yet, get the weighted mean value of all gray-scale values on histogram as threshold value, the value of N is determined by experiment.
Preferably, the described method to the smoothing process of histogram data is: to the histogrammic value of left and right 3 points of each point in histogram X-axis except first and last point average level and smooth as this point after histogrammic value.The histogrammic value of first and second point average level and smooth as first after histogrammic value.The histogrammic value of point second from the bottom and last point average level and smooth as last point after histogrammic value.
Further, the gray-scale value of pixel after step 1 processes only has 0 and 255 two kind, gray-scale value be 0 pixel belong to edge of crack point, gray-scale value be 255 belong to non-edge of crack point.
Further, extract two curves of edge of crack described in step 2 and determine that the method for fracture strike comprises the following steps:
(1) line by line scan by ordinate order from small to large, till scanning and having 2 gray-scale values to be the pixel of 0 on a same row.
(2) take 2 pixels in step (1) as the starting point of two curves, utilize the path finding algorithm in the anti-game of tower, calculate 2 curves of edge of crack.
(3) whether 2 curves that determining step (2) obtains there is the point of repetition, if there is the point of repetition, then 2 curves are same curves, scanning ordinate value is added 1, repeat step (1), (2), (3), until find 2 different curves.If failure, then perform step (4).
(4) scan by column by horizontal ordinate order from small to large, on same row, have 2 gray-scale values to be the pixel of 0 until scan.
(5) take 2 pixels in step (4) as the starting point of two curves, utilize the path finding algorithm in the anti-game of tower, calculate 2 curves of edge of crack.
(6) whether 2 curves that determining step (5) obtains there is the point of repetition, if there is the point of repetition, then 2 curves are same curves, scanning abscissa value is added 1, repeat step (4), (5), (6), until find 2 different curves.
The computing method of described 2 boundary curves are identical, and the method calculating a curve is as follows:
First scanning pixel is out added in a search queue, then iteration is carried out, from search queue, take out a pixel at every turn, to check around it the gray-scale value of 8 adjacent pixels, by gray-scale value be 0 pixel join in described search queue, until searched for last pixel in queue.Preserving the gray-scale value searched out in iterative process is the pixel of 0, and records a upper pixel of current pixel point and the distance between current pixel point and starting point.
After having searched for, to find with start point distance from maximum pixel, i.e. terminal, calculated apart from the path between maximum pixel and starting point by recursive algorithm, thus obtain a boundary curve in crack.
Described curve is exactly the trend in crack from the direction of origin-to-destination.
Further, the distance between described current pixel point and starting point equals the number of the pixel that the path from starting point to current pixel point comprises, and computing method are that a upper distance between pixel and starting point adds 1.
Further, as follows to the method for every bar curve filtering described in step 3:
(1) first and last two pixel of junction curve obtains a straight line, obtains the distance of all the other each pixels to this straight line.
(2) if the maximal value of the distance of trying to achieve in step (1) be less than or equal to limit poor, then delete all pixels between first and last two pixel on described straight line; If it is poor that the maximal value of described distance is greater than limit, then retain the pixel that described ultimate range is corresponding, and with this pixel for boundary is divided into two parts curve, step (1), (2) are repeated to this two parts curve, until no longer occur that the maximal value of described distance is less than or equal to the pixel of limit difference.The value of described limit difference is determined by experiment.
After filtering process, described curve splits into many line segments, and obtain the normal distance between every article of line segment on the 1st article of curve and every article of line segment on the 2nd article of curve respectively, the maximal value of described normal distance is fracture width.
Compared with prior art, the present invention has following beneficial effect:
(1) the present invention adopts image processing techniques to carry out the extraction of body surface edge of crack curve, does not need the trend manually adjusting crack, as long as crack occurs in the picture, and just can the automatic trend of crack identification and the breadth extreme in crack.
(2) the present invention adopts and carries out image procossing based on histogrammic image overall binarization method, can retain the pixel of edge of crack to greatest extent, improve measuring accuracy.
(3) two curves at the method for the invention first fracture edge before calculating fracture width carry out filtering process, the boundary curve in crack not only can be made more level and smooth, and decrease the quantity that curve is put, improve the computing velocity of fracture width.The interpretation time of the method for the invention fracture width is less than 1/10 of semi-automatic method of identification, is less than 1/30 of artificial visually examine's method.
Accompanying drawing explanation
Fig. 1 is the method flow diagram of the automatic recognition object surface crack trend that relates to of the embodiment of the present invention and width;
Fig. 2 is the process flow diagram of image processing method in Fig. 1;
Fig. 3 is the process flow diagram calculating overall binary-state threshold;
Fig. 4 is the process flow diagram of calculating one crack boundary curve.
Embodiment
Below in conjunction with drawings and Examples, the present invention will be further described.
The present invention proposes a kind of method of automatic recognition object surface crack trend and width, and process flow diagram as shown in figures 1-4, said method comprising the steps of:
Step 1, carries out gray processing, binaryzation and filtering process to image, then carries out Canny rim detection, obtain the image of edge of crack pixel.
Binary conversion treatment adopts a kind of based on histogrammic image overall binarization method, comprises the following steps:
Step 1.1, according to the gray-scale value compute histograms obtained after image gray processing process.
Step 1.2, if described histogram is a bimodal histogram, gets two peak-to-peak the lowest point values as threshold value; If described histogram is not a bimodal histogram, then to the smoothing process of histogram data, method is as follows:
To the histogrammic value of left and right 3 points of each point in histogram X-axis except first and last point average level and smooth as this point after histogrammic value.The histogrammic value of first and second point average level and smooth as first after histogrammic value.The histogrammic value of point second from the bottom and last point average level and smooth as last point after histogrammic value.
Step 1.3, repeats step 1.2, until described histogram becomes a bimodal histogram.
If do not obtain a bimodal histogram after being repeated N time yet, get the weighted mean value of all gray-scale values on histogram as threshold value.The value of N is determined by experiment, and the present embodiment N gets 1000.
The gray-scale value of the pixel after above-mentioned process only has 0 and 255 two kind, gray-scale value be 0 pixel belong to edge of crack point, gray-scale value be 255 belong to non-edge of crack point.
Step 2, extracts two curves of edge of crack, and determines the trend in crack based on the path finding algorithm in the anti-game of tower.
Step 2.1, lines by line scan by ordinate order from small to large, till scanning and having 2 gray-scale values to be the pixel of 0 on a same row.
Step 2.2, take the pixel of 2 in step 2.1 as the starting point of 2 curves, utilizes the path finding algorithm in the anti-game of tower, calculates 2 curves of edge of crack.
Article 2, the computing method of boundary curve are identical, and the method calculating a curve is as follows:
First scanning pixel is out added in a search queue, then iteration is carried out, from search queue, take out a pixel at every turn, to check around it the gray-scale value of 8 adjacent pixels, by gray-scale value be 0 pixel join in described search queue, until searched for last pixel in queue.Preserving the gray-scale value searched out in iterative process is the pixel of 0, and record a upper pixel of current pixel point and the distance between current pixel point and starting point (distance between current pixel point and starting point refers to the number of the pixel that starting point to the path of current pixel point comprises, and computing method are that a upper distance between pixel and starting point adds 1).
After having searched for, find with start point distance from maximum pixel, saved a upper pixel of each pixel above, calculated apart from the path between maximum pixel and starting point by recursive algorithm, so this paths is exactly a boundary curve in crack.This curve is exactly the trend in crack from the direction of origin-to-destination (pixel apart from maximum).
Step 2.3, whether 2 curves that determining step 2.2 obtains has the point of repetition, if there is the point of repetition, just thinks same curve, scanning ordinate value is added 1, repeats step 2.1, step 2.2, step 2.3, until find 2 different curves.If failure, then perform step 2.4.
Step 2.4, scans by column by horizontal ordinate order from small to large, has 2 gray-scale values to be the pixel of 0 until scan on same row.
Step 2.5, take the pixel of 2 in step 2.4 as the starting point of two curves, utilizes the path finding algorithm in the anti-game of tower, and calculate 2 curves of edge of crack, computing method are with step 2.2.
Step 2.6, whether 2 curves that determining step 2.5 obtains has the point of repetition, if there is the point of repetition, then 2 curves are same curves, scanning abscissa value is added 1, repeats step 2.4, step 2.5, step 2.6, until find 2 different curves.If failure, then function returns failure.
Step 3, carries out filtering process to two curves of the edge of crack that step 2 obtains, and calculates fracture width.
Processing speed is improved, the smoothing filtering process of 2 curves at first fracture edge before calculating fracture width in order to reduce calculated amount, as follows to the method for every bar curve filtering:
Step 3.1, first and last two pixel of junction curve obtains a straight line, obtains the distance of all the other each pixels to this straight line.
Step 3.2, if the maximal value of the distance of trying to achieve in step 3.1 be less than or equal to limit poor, then delete all pixels between first and last two pixel on described straight line; If it is poor that the maximal value of described distance is greater than limit, then retain the pixel that described ultimate range is corresponding, and with this pixel for boundary is divided into two parts curve, step 3.1,3.2 is repeated to this two parts curve, until no longer occur that the maximal value of described distance is less than or equal to the pixel of limit difference.
The value of described limit difference is determined by experiment.In order to ensure level and smooth after curve shape and smoothly lastly to cause, reduce measuring error, described in the present embodiment, limit differs from and gets 1.After curve filtering completes, curve becomes and is made up of many line segments, and the width in crack is exactly the maximal value of normal direction distance between all line segments on the boundary curve of 2, crack.
Table 1 gives and adopts the method for the invention and semi-automatic method of identification of the prior art and artificial visually examine's method, carries out the Contrast on effect of fracture strike and width identification.As shown in Table 1, the method of the invention is not only better than semi-automatic method of identification and artificial visually examine's method in measuring accuracy, and in processing speed, being obviously better than other two kinds of methods, the interpretation time of fracture width is less than 1/10 of semi-automatic method of identification, 1/30 of artificial visually examine's method.In addition, the method for the invention can be moved towards by crack identification automatically, and fracture trend does not require.
The contrast of table 1 the method for the invention and prior art
The inventive method | Semi-automatic method of identification | Artificial visually examine's method | |
Measuring accuracy | ≤0.005mm | ≥0.005mm | ≥0.05mm |
The fracture width interpretation time | ≤ 1 second | >=10 seconds | >=30 seconds |
Whether support automatic discrimination | Support | Support | Do not support |
Whether in real time interpretation | Support | Part is supported | Do not support |
Whether fracture strike has requirement | No | Horizontal or vertical direction | Vertical direction |
Fracture strike can be identified | Energy | Can not | Can not |
The invention is not restricted to above-mentioned embodiment, those skilled in the art make to any apparent improvement of above-mentioned embodiment or change, all can not exceed the protection domain of design of the present invention and claims.
Claims (7)
1. a method for automatic recognition object surface crack trend and width, is characterized in that comprising the following steps:
Step 1, carries out gray processing, binaryzation and filtering process to image, then carries out Canny rim detection, obtain the image of edge of crack pixel;
Step 2, extracts two curves of edge of crack, and determines the trend in crack;
Step 3, carries out filtering process to two curves of the edge of crack that step 2 obtains, and calculates fracture width.
2. the method for automatic recognition object surface crack trend according to claim 1 and width, it is characterized in that, the image binaryzation process described in step 1 adopts a kind of based on histogrammic image overall binarization method, comprises the following steps:
(1) according to the gray-scale value compute histograms obtained after image gray processing process;
(2) if described histogram is a bimodal histogram, two peak-to-peak the lowest point values are got as threshold value; If described histogram is not a bimodal histogram, then to the smoothing process of histogram data;
(3) step (2) is repeated, until described histogram becomes a bimodal histogram; If do not obtain a bimodal histogram after being repeated N time yet, get the weighted mean value of all gray-scale values on histogram as threshold value; The value of N is determined by experiment.
3. the method for automatic recognition object surface crack trend according to claim 2 and width, it is characterized in that, the described method to the smoothing process of histogram data is: to the histogrammic value of left and right 3 points of each point in histogram X-axis except first and last point average level and smooth as this point after histogrammic value; The histogrammic value of first and second point average level and smooth as first after histogrammic value; The histogrammic value of point second from the bottom and last point average level and smooth as last point after histogrammic value.
4. the method for automatic recognition object surface crack trend according to claim 1 and width, it is characterized in that, the gray-scale value of pixel after step 1 processes only has 0 and 255 two kind, gray-scale value be 0 pixel belong to edge of crack point, gray-scale value be 255 belong to non-edge of crack point.
5. the method for automatic recognition object surface crack trend according to claim 1 and width, is characterized in that, extract two curves of edge of crack and determine that the method for fracture strike comprises the following steps described in step 2:
(1) line by line scan by ordinate order from small to large, till scanning and having 2 gray-scale values to be the pixel of 0 on a same row;
(2) take 2 pixels in step (1) as the starting point of two curves, utilize the path finding algorithm in the anti-game of tower, calculate 2 curves of edge of crack;
(3) whether 2 curves that determining step (2) obtains there is the point of repetition, if there is the point of repetition, then 2 curves are same curves, scanning ordinate value is added 1, repeat step (1), (2), (3), until find 2 different curves; If failure, then perform step (4);
(4) scan by column by horizontal ordinate order from small to large, on same row, have 2 gray-scale values to be the pixel of 0 until scan;
(5) take 2 pixels in step (4) as the starting point of two curves, utilize the path finding algorithm in the anti-game of tower, calculate 2 curves of edge of crack;
(6) whether 2 curves that determining step (5) obtains there is the point of repetition, if there is the point of repetition, then 2 curves are same curves, scanning abscissa value is added 1, repeat step (4), (5), (6), until find 2 different curves;
The computing method of described 2 boundary curves are identical, and the method calculating a curve is as follows:
First scanning pixel is out added in a search queue, then iteration is carried out, from search queue, take out a pixel at every turn, to check around it the gray-scale value of 8 adjacent pixels, by gray-scale value be 0 pixel join in described search queue, until searched for last pixel in queue; Preserving the gray-scale value searched out in iterative process is the pixel of 0, and records a upper pixel of current pixel point and the distance between current pixel point and starting point;
After having searched for, to find with start point distance from maximum pixel, i.e. terminal, calculated apart from the path between maximum pixel and starting point by recursive algorithm, thus obtain a boundary curve in crack;
Described curve is exactly the trend in crack from the direction of origin-to-destination.
6. the method for automatic recognition object surface crack trend according to claim 5 and width, it is characterized in that, distance between described current pixel point and starting point equals the number of the pixel that the path from starting point to current pixel point comprises, and computing method are that a upper distance between pixel and starting point adds 1.
7. the method for automatic recognition object surface crack trend according to claim 1 and width, is characterized in that, as follows to the method for every bar curve filtering described in step 3:
(1) first and last two pixel of junction curve obtains a straight line, obtains the distance of all the other each pixels to this straight line;
(2) if the maximal value of the distance of trying to achieve in step (1) be less than or equal to limit poor, then delete all pixels between first and last two pixel on described straight line; If it is poor that the maximal value of described distance is greater than limit, then retain the pixel that described ultimate range is corresponding, and with this pixel for boundary is divided into two parts curve, step (1), (2) are repeated to this two parts curve, until no longer occur that the maximal value of described distance is less than or equal to the pixel of limit difference; The value of described limit difference is determined by experiment;
After filtering process, described curve splits into many line segments, and obtain the normal distance between every article of line segment on the 1st article of curve and every article of line segment on the 2nd article of curve respectively, the maximal value of described normal distance is fracture width.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510661260.8A CN105389793B (en) | 2015-10-14 | 2015-10-14 | A kind of method of automatic identification body surface fracture strike and width |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510661260.8A CN105389793B (en) | 2015-10-14 | 2015-10-14 | A kind of method of automatic identification body surface fracture strike and width |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105389793A true CN105389793A (en) | 2016-03-09 |
CN105389793B CN105389793B (en) | 2018-10-19 |
Family
ID=55422043
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510661260.8A Active CN105389793B (en) | 2015-10-14 | 2015-10-14 | A kind of method of automatic identification body surface fracture strike and width |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105389793B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108919377A (en) * | 2018-07-28 | 2018-11-30 | 林光琴 | A kind of geotechnical engineering wall-rock crack detection system |
CN109448012A (en) * | 2018-10-29 | 2019-03-08 | 山东浪潮云信息技术有限公司 | A kind of method for detecting image edge and device |
CN110207592A (en) * | 2019-04-15 | 2019-09-06 | 深圳高速工程检测有限公司 | Building cracks measurement method, device, computer equipment and storage medium |
CN113689453A (en) * | 2021-08-24 | 2021-11-23 | 中石化石油工程技术服务有限公司 | Method, device and equipment for automatically identifying well logging image cracks and storage medium |
CN114111602A (en) * | 2021-11-22 | 2022-03-01 | 招商局重庆交通科研设计院有限公司 | Bridge surface crack width calculation method based on image technology |
CN117132602A (en) * | 2023-10-27 | 2023-11-28 | 湖南三昌泵业有限公司 | Visual inspection method for defects of centrifugal pump impeller |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1563891A (en) * | 2004-04-20 | 2005-01-12 | 长安大学 | System and method for discriminating road gap |
CN101413901A (en) * | 2008-12-01 | 2009-04-22 | 南京航空航天大学 | Surface fatigue crack detecting method based on CCD image characteristic |
-
2015
- 2015-10-14 CN CN201510661260.8A patent/CN105389793B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1563891A (en) * | 2004-04-20 | 2005-01-12 | 长安大学 | System and method for discriminating road gap |
CN101413901A (en) * | 2008-12-01 | 2009-04-22 | 南京航空航天大学 | Surface fatigue crack detecting method based on CCD image characteristic |
Non-Patent Citations (6)
Title |
---|
BARRY WILKINSON等: "《并行程序设计(第二版)》", 31 May 2005 * |
叶贵如等: "基于数字图像处理的表面裂缝宽度测量", 《公路交通科技》 * |
吴丹等: "基于直方图分析和OTSU算法的文字图像二值化", 《计算机与现代化》 * |
张宏等: "《地理信息***算法基础》", 30 June 2006 * |
徐志刚等: "基于直方图估计和形状分析的沥青路面裂缝识别算法", 《仪器仪表学报》 * |
邹小平等: "《纳米材料与敏化太阳电池》", 31 December 2014 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108919377A (en) * | 2018-07-28 | 2018-11-30 | 林光琴 | A kind of geotechnical engineering wall-rock crack detection system |
CN108919377B (en) * | 2018-07-28 | 2020-07-17 | 嘉兴麦瑞网络科技有限公司 | Geotechnical engineering surrounding rock crack detection system |
CN109448012A (en) * | 2018-10-29 | 2019-03-08 | 山东浪潮云信息技术有限公司 | A kind of method for detecting image edge and device |
CN110207592A (en) * | 2019-04-15 | 2019-09-06 | 深圳高速工程检测有限公司 | Building cracks measurement method, device, computer equipment and storage medium |
CN110207592B (en) * | 2019-04-15 | 2021-11-09 | 深圳高速工程检测有限公司 | Building crack measuring method and device, computer equipment and storage medium |
CN113689453A (en) * | 2021-08-24 | 2021-11-23 | 中石化石油工程技术服务有限公司 | Method, device and equipment for automatically identifying well logging image cracks and storage medium |
CN114111602A (en) * | 2021-11-22 | 2022-03-01 | 招商局重庆交通科研设计院有限公司 | Bridge surface crack width calculation method based on image technology |
CN117132602A (en) * | 2023-10-27 | 2023-11-28 | 湖南三昌泵业有限公司 | Visual inspection method for defects of centrifugal pump impeller |
CN117132602B (en) * | 2023-10-27 | 2024-01-02 | 湖南三昌泵业有限公司 | Visual inspection method for defects of centrifugal pump impeller |
Also Published As
Publication number | Publication date |
---|---|
CN105389793B (en) | 2018-10-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105389793A (en) | Method for automatically identifying trend and width of fracture in object surface | |
CN109377485B (en) | Machine vision detection method for instant noodle packaging defects | |
CN108038883B (en) | Crack detection and identification method applied to highway pavement video image | |
CN104990925B (en) | One kind is based on gradient multi thresholds optimization defect inspection method | |
CN104268872B (en) | Consistency-based edge detection method | |
CN113109368B (en) | Glass crack detection method, device, equipment and medium | |
CN109540925B (en) | Complex ceramic tile surface defect detection method based on difference method and local variance measurement operator | |
CN109376740A (en) | A kind of water gauge reading detection method based on video | |
CN110555382A (en) | Finger vein identification method based on deep learning and Wasserstein distance measurement | |
CN113469990B (en) | Pavement disease detection method and device | |
CN110033458A (en) | It is a kind of based on pixel gradient distribution image threshold determine method | |
JP6338429B2 (en) | Subject detection apparatus, subject detection method, and program | |
CN110717900A (en) | Pantograph abrasion detection method based on improved Canny edge detection algorithm | |
CN113487563B (en) | EL image-based self-adaptive detection method for hidden cracks of photovoltaic module | |
CN111476804A (en) | Method, device and equipment for efficiently segmenting carrier roller image and storage medium | |
CN112085699B (en) | Pavement crack extraction method based on two-dimensional image | |
CN111429372A (en) | Method for enhancing edge detection effect of low-contrast image | |
CN114332079A (en) | Plastic lunch box crack detection method, device and medium based on image processing | |
CN104036514A (en) | Circle detection method based on histogram peak value search | |
CN116740072A (en) | Road surface defect detection method and system based on machine vision | |
CN109636822B (en) | Improved Canny self-adaptive edge extraction method based on newly-constructed membership function | |
CN109671084B (en) | Method for measuring shape of workpiece | |
CN117893550A (en) | Moving object segmentation method under complex background based on scene simulation | |
CN113781413A (en) | Electrolytic capacitor positioning method based on Hough gradient method | |
CN117853510A (en) | Canny edge detection method based on bilateral filtering and self-adaptive threshold |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |