CN103942809A - Method for detecting joint fissures in rock images - Google Patents
Method for detecting joint fissures in rock images Download PDFInfo
- Publication number
- CN103942809A CN103942809A CN201410197540.3A CN201410197540A CN103942809A CN 103942809 A CN103942809 A CN 103942809A CN 201410197540 A CN201410197540 A CN 201410197540A CN 103942809 A CN103942809 A CN 103942809A
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- straight line
- joint fissure
- line
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of rock joint fissure detection, in particular to a method for detecting joint fissures in rock images. The method comprises a first step of performing denoising and enhancement processing on binary images of rock images and performing skeleton extraction; a second step of adopting Hough transformation to respectively detect main lines (connected domains) of every joint fissure section in the binary images having undergone the processing in the first step; and a third step of adopting a Bresenham algorithm to perform segmented expansion on every main line and obtaining joint fissure areas. The method improves detecting effects of the rock joint fissures in the images.
Description
Technical field
The present invention relates to detection technique field, rock joint crack, particularly a kind of method that detects joint fissure in rock image.
Background technology
In the processing of rock joint crack image, joint fissure, as important target, carries out automatic Study of recognition to it and has great importance.Because the development degree of joint fissure is all significant for design and the safe early warning of many rock engineerings.Therefore, rock joint crack be automatically identified as problem anxious to be resolved.
Since the nineties in last century, people have just started to pay close attention to the detection identification of joint fissure image, have also proposed a lot of methods, as by obtaining image with different sensors, go out joint fissure target from region of interesting extraction, based on the target detection of rock engineering knowledge, contextual information.In recent years, a large amount of researchers has produced very large interest to the automatic identification of joint fissure again, and has proposed some recognition methodss for some data sources.
Summary of the invention
The object of the present invention is to provide a kind of method that detects joint fissure in rock image, the method has improved the detection effect of joint fissure in rock image.
For achieving the above object, technical scheme of the present invention is: a kind of method that detects joint fissure in rock image, comprises the following steps:
Step 1: the bianry image of rock image is carried out to denoising and hole filling;
Step 2: adopt Hough to convert the main line that detects respectively each joint fissure section in step 1 bianry image after treatment, i.e. connected region;
Step 3: adopt Bresenham algorithm to expand every main line segmentation, obtain joint fissure region.
Further, in step 1, utilize mathematical morphology and logical operation to carry out denoising and hole filling to described bianry image, comprise the following steps:
Step 1.1: the bianry image of described rock image is carried out to opening operation, grain noise is removed from image;
Step 1.2: extract separately independently connected region from step 1.1 image after treatment, and give respectively mark value, then the each connected region of searching loop calculate its area, if area is less than the first threshold of setting, judge that this connected region is not target area, it is 0 that all this connected region pixel values are composed, and to eliminate little noise particles, leaves the target area of macrostructure;
Step 1.3: step 1.2 image after treatment is carried out to hole filling;
Step 1.4: joint fissure is carried out to main line extraction, and concrete grammar is as follows:
1) based on Otsu threshold method, image is divided into first area and second area;
2) respectively circular treatment is carried out in first area and second area, if the pixel p in first area meets the condition of following G1, G2 and G3 simultaneously, just remove described pixel p, assignment is 0; If the pixel p in second area meets the condition of following G1, G2 and G3 simultaneously, just remove described pixel p;
Condition G1: when tested measuring point p meets
time, wherein
, x
1, x
2..., x
88 neighborhoods of pixel p,
x h (
p) factor of expression;
Condition G2: when tested measuring point p meets
,
time, wherein
,
n 1(
p),
n 2(
p) represent respectively two different judgement parameters;
Condition G3:
,
Finally obtain the bianry image after denoising and hole filling treatment.
Further, in step 1.2, from step 1.1 image after treatment, extract separately as follows independently connected region: image is carried out to point by point scanning, if current pixel value is 0, just move on to next scanning position; If current pixel value is 1, check its 8 neighbors, until 8 adjacent pixel values are all 0, judge that this is as a connected region.
Further, in step 2, adopt Hough conversion segmentation to detect the main line of each joint fissure, to connect interrupted joint fissure, comprise the following steps:
Step 2.1: establishing y=k*x+b is the straight line in rock image x-y plane, and wherein k and b are parameters, represent respectively slope and intercept; Cross a bit (x
0, y
0) the parameter of all straight lines all can meet equation y
0=kx
0+ b, i.e. point (x
0, y
0) determine cluster straight line, and equation y
0=kx
0+ b is straight line in parameter k-b plane;
Step 2.2: a foreground pixel point in rock image x-y plane, joint fissure is put the straight line in corresponding parameter k-b plane, is mapped as the line of concurrent in parameter k-b plane by the point of conllinear in rock image x-y plane;
Step 2.3: the multiple little line segment generating after processing for step 2.1 and step 2.2, will be positioned at collinear line segment and merge, in the line segment after merging, length be greater than setting Second Threshold be the main line detecting.
Further, in step 3, to every main line segmentation expansion, determine joint fissure region as follows:
The mesh lines of constructing virtual on image, each grid represents a pixel; Select a direction to the straight line on image, by the intersection point of finding out this straight line and each mesh lines from the order of origin-to-destination, find out the next pixel nearest with each intersection point, these pixels couple together generate straight-line segment, be the straight line nearest with initial straight line or broken line;
If straight line is y=kx+b, m=△ y/ △ x, △ y, △ x represent respectively the increment of y, x direction, the pixel of straight line can only round numerical coordinates; Suppose that again on straight line, i pixel coordinate is (x
i, y
i), it is Points on Straight Line (x
i, y
i) optimal approximation, and x
i=x
i(establishing m<1), so, on straight line, the possible position of next pixel is (x
i+1, y
i) or (x
i+1, y
i+1); Work as x=x
i+1time, the y value of Points on Straight Line is y=m (x
i+1)+b, this point is from pixel (x
i+1, y
i) and pixel (x
i+1, y
i+1) distance be respectively d1 and d2, select as follows next pixel:
(1) d1>d2, illustrates on straight line that mathematical point is from (x
i+1, y
i+1) pixel is nearer, next pixel is got (x
i+1, y
i+1);
(2) d1<d2, illustrates on straight line that mathematical point is from (x
i+1, y
i) pixel is nearer, next pixel is got (x
i+1, y
i);
(3) d1=d2, illustrates on straight line that mathematical point is from (x
i+1, y
i), (x
i+1, y
i+1) distance of two pixels equates, appoints that to get be wherein next pixel;
Thereby the show line that extends out that obtains every straight line or broken line, finally obtains joint fissure region.
The invention has the beneficial effects as follows effectively overcome when joint fissure region longer and narrower, when particularly joint fissure width only has 2～3 pixels, the bad shortcoming of joint fissure effect of utilizing the method for marginal information detection of straight lines to detect in prior art.The method is not to detect the parallel lines pair at joint fissure edge, but according to the whole target area of the axis detection of joint fissure, has good effect for the joint fissure extracting in geologic image, has application prospect very widely.
Brief description of the drawings
Fig. 1 is the realization flow figure of the embodiment of the present invention.
Fig. 2 is that the embodiment of the present invention is determined a coordinate plane schematic diagram corresponding in the process of joint fissure region.
Fig. 3 is that the embodiment of the present invention is determined another coordinate plane schematic diagram corresponding in the process of joint fissure region.
Embodiment
Below in conjunction with drawings and the specific embodiments, the present invention is described in further detail.
The present invention detects the method for joint fissure in rock image, as shown in Figure 1, comprises the following steps:
Step 1: rock image is converted into bianry image, the bianry image of rock image is carried out to denoising and hole filling.
In step 1, utilize mathematical morphology and logical operation to carry out denoising and hole filling to described bianry image, comprise the following steps:
Step 1.1: the bianry image of described rock image is carried out to opening operation, grain noise is removed from image, more interrupted joint fissure is sewed up.
Step 1.2: extract separately independently connected region from step 1.1 image after treatment, and give respectively mark value, then the each connected region of searching loop calculate its area, if area is less than the first threshold of setting, judge that this connected region is not target area, it is 0 that all this connected region pixel values are composed, and to eliminate little noise particles, leaves the target area of macrostructure.
In step 1.2, from step 1.1 image after treatment, extract separately as follows independently connected region: image is carried out to point by point scanning, if current pixel value is 0, just move on to next scanning position; If current pixel value is 1, check its 8 neighbors, until 8 adjacent pixel values are all 0, judge that this is as a connected region.
Step 1.3: in order to reduce the extraction of hole to joint fissure skeleton, step 1.2 image after treatment is carried out to hole filling.
Step 1.4: joint fissure is carried out to main line extraction, and concrete grammar is as follows:
1) based on Otsu threshold method, image is divided into first area and second area;
2) respectively circular treatment is carried out in first area and second area, if the pixel p in first area meets the condition of following G1, G2 and G3 simultaneously, just remove described pixel p (assignment is 0); If the pixel p in second area meets the condition of following G1, G2 and G3 simultaneously, just remove described pixel p;
Condition G1: when tested measuring point p meets
time, wherein
, x
1, x
2..., x
88 neighborhoods of pixel p,
x h (
p) factor of expression;
Condition G2: when tested measuring point p meets
,
time, wherein
,
n 1(
p),
n 2(
p) represent respectively two different judgement parameters;
Condition G3:
,
Finally obtain the bianry image after denoising and hole filling treatment.
Step 2: adopt Hough to convert the main line that detects respectively each joint fissure section in step 1 bianry image after treatment, i.e. connected region.
In step 2, adopt Hough conversion segmentation to detect the main line of each joint fissure, to connect interrupted joint fissure, comprise the following steps:
Step 2.1: establishing y=k*x+b is the straight line in rock image x-y plane, and wherein k and b are parameters, represent respectively slope and intercept; Cross a bit (x
0, y
0) the parameter of all straight lines all can meet equation y
0=kx
0+ b, i.e. point (x
0, y
0) determine cluster straight line, and equation y
0=kx
0+ b is straight line in parameter k-b plane;
Step 2.2: a foreground pixel point in rock image x-y plane, joint fissure is put the straight line in corresponding parameter k-b plane, is mapped as the line of concurrent in parameter k-b plane by the point of conllinear in rock image x-y plane;
Step 2.3: the multiple little line segment generating after processing for step 2.1 and step 2.2, will be positioned at collinear line segment and merge, in the line segment after merging, length be greater than setting Second Threshold be the main line detecting.
Step 3: adopt Bresenham algorithm to expand every main line segmentation, obtain joint fissure region.
In step 3, to every main line segmentation expansion, determine joint fissure region as follows:
The mesh lines of constructing virtual on image, each grid represents a pixel; Select a direction to the straight line on image, by the intersection point of finding out this straight line and each mesh lines from the order of origin-to-destination, find out the next pixel nearest with each intersection point, these pixels couple together generate straight-line segment, be the straight line nearest with initial straight line or broken line;
If straight line is y=kx+b, m=△ y/ △ x, △ y, △ x represent respectively the increment of y, x direction, from Fig. 2,3, can find out, the pixel of straight line can only round numerical coordinates; Suppose that again on straight line, i pixel coordinate is (x
i, y
i), according to the concept of discrete mathematics, it is Points on Straight Line (x
i, y
i) optimal approximation, and x
i=x
i(establishing m<1), so, on straight line, the possible position of next pixel is (x
i+1, y
i) or (x
i+1, y
i+1); Work as x=x
i+1time, the y value of Points on Straight Line is y=m (x
i+1)+b, this point is from pixel (x
i+1, y
i) and pixel (x
i+1, y
i+1) distance be respectively d1 and d2, select as follows next pixel:
(1) d1>d2, illustrates on straight line that mathematical point is from (x
i+1, y
i+1) pixel is nearer, next pixel is got (x
i+1, y
i+1);
(2) d1<d2, illustrates on straight line that mathematical point is from (x
i+1, y
i) pixel is nearer, next pixel is got (x
i+1, y
i);
(3) d1=d2, illustrates on straight line that mathematical point is from (x
i+1, y
i), (x
i+1, y
i+1) distance of two pixels equates, appoints that to get be wherein next pixel;
Thereby the show line that extends out that obtains every straight line or broken line, finally obtains joint fissure region.
Be more than preferred embodiment of the present invention, all changes of doing according to technical solution of the present invention, when the function producing does not exceed the scope of technical solution of the present invention, all belong to protection scope of the present invention.
Claims (5)
1. a method that detects joint fissure in rock image, is characterized in that, comprises the following steps:
Step 1: the bianry image of rock image is carried out to denoising and hole filling;
Step 2: adopt Hough to convert the main line that detects respectively each joint fissure section in step 1 bianry image after treatment, i.e. connected region;
Step 3: adopt Bresenham algorithm to expand every main line segmentation, obtain joint fissure region.
2. the method for joint fissure in detection rock image according to claim 1, is characterized in that, in step 1, utilizes mathematical morphology and logical operation to carry out denoising and hole filling to described bianry image, comprises the following steps:
Step 1.1: the bianry image of described rock image is carried out to opening operation, grain noise is removed from image;
Step 1.2: extract separately independently connected region from step 1.1 image after treatment, and give respectively mark value, then the each connected region of searching loop calculate its area, if area is less than the first threshold of setting, judge that this connected region is not target area, it is 0 that all this connected region pixel values are composed, and to eliminate little noise particles, leaves the target area of macrostructure;
Step 1.3: step 1.2 image after treatment is carried out to hole filling;
Step 1.4: joint fissure is carried out to main line extraction, and concrete grammar is as follows:
1) based on Otsu threshold method, image is divided into first area and second area;
2) respectively circular treatment is carried out in first area and second area, if the pixel p in first area meets the condition of following G1, G2 and G3 simultaneously, just remove described pixel p, assignment is 0; If the pixel p in second area meets the condition of following G1, G2 and G3 simultaneously, just remove described pixel p;
Condition G1: when tested measuring point p meets
time, wherein
, x
1, x
2..., x
88 neighborhoods of pixel p,
x h (
p) factor of expression;
Condition G2: when tested measuring point p meets
,
time, wherein
,
n 1(
p),
n 2(
p) represent respectively two different judgement parameters;
Condition G3:
,
Finally obtain the bianry image after denoising and hole filling treatment.
3. the method for joint fissure in detection rock image according to claim 2, it is characterized in that, in step 1.2, from step 1.1 image after treatment, extract separately as follows independently connected region: image is carried out to point by point scanning, if current pixel value is 0, just move on to next scanning position; If current pixel value is 1, check its 8 neighbors, until 8 adjacent pixel values are all 0, judge that this is as a connected region.
4. the method for joint fissure in detection rock image according to claim 2, is characterized in that: in step 2, adopt Hough conversion segmentation to detect the main line of each joint fissure, i.e. connected region, to connect interrupted joint fissure, comprises the following steps:
Step 2.1: establishing y=k*x+b is the straight line in rock image x-y plane, and wherein k and b are parameters, represent respectively slope and intercept; Cross a bit (x
0, y
0) the parameter of all straight lines all can meet equation y
0=kx
0+ b, i.e. point (x
0, y
0) determine cluster straight line, and equation y
0=kx
0+ b is straight line in parameter k-b plane;
Step 2.2: a foreground pixel point in rock image x-y plane, joint fissure is put the straight line in corresponding parameter k-b plane, is mapped as the line of concurrent in parameter k-b plane by the point of conllinear in rock image x-y plane;
Step 2.3: the multiple little line segment generating after processing for step 2.1 and step 2.2, will be positioned at collinear line segment and merge, in the line segment after merging, length be greater than setting Second Threshold be the main line detecting.
5. the method for joint fissure in detection rock image according to claim 4, is characterized in that: in step 3, to every main line segmentation expansion, determine joint fissure region as follows:
The mesh lines of constructing virtual on image, each grid represents a pixel; Select a direction to the straight line on image, by the intersection point of finding out this straight line and each mesh lines from the order of origin-to-destination, find out the next pixel nearest with each intersection point, these pixels couple together generate straight-line segment, be the straight line nearest with initial straight line or broken line;
If straight line is y=kx+b, m=△ y/ △ x, △ y, △ x represent respectively the increment of y, x direction, the pixel of straight line can only round numerical coordinates; Suppose that again on straight line, i pixel coordinate is (x
i, y
i), it is Points on Straight Line (x
i, y
i) optimal approximation, and x
i=x
i(establishing m<1), so, on straight line, the possible position of next pixel is (x
i+1, y
i) or (x
i+1, y
i+1); Work as x=x
i+1time, the y value of Points on Straight Line is y=m (x
i+1)+b, this point is from pixel (x
i+1, y
i) and pixel (x
i+1, y
i+1) distance be respectively d1 and d2, select as follows next pixel:
(1) d1>d2, illustrates on straight line that mathematical point is from (x
i+1, y
i+1) pixel is nearer, next pixel is got (x
i+1, y
i+1);
(2) d1<d2, illustrates on straight line that mathematical point is from (x
i+1, y
i) pixel is nearer, next pixel is got (x
i+1, y
i);
(3) d1=d2, illustrates on straight line that mathematical point is from (x
i+1, y
i), (x
i+1, y
i+1) distance of two pixels equates, appoints that to get be wherein next pixel;
Thereby the show line that extends out that obtains every straight line or broken line, finally obtains joint fissure region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410197540.3A CN103942809A (en) | 2014-05-12 | 2014-05-12 | Method for detecting joint fissures in rock images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410197540.3A CN103942809A (en) | 2014-05-12 | 2014-05-12 | Method for detecting joint fissures in rock images |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103942809A true CN103942809A (en) | 2014-07-23 |
Family
ID=51190460
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410197540.3A Pending CN103942809A (en) | 2014-05-12 | 2014-05-12 | Method for detecting joint fissures in rock images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103942809A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106023197A (en) * | 2016-05-18 | 2016-10-12 | 南京师范大学 | Automated identification and extraction method of vertical stratum |
| CN106683060A (en) * | 2017-01-03 | 2017-05-17 | 北京大学 | Firing model removing microwave noise method switch |
| CN107680092A (en) * | 2017-10-12 | 2018-02-09 | 中科视拓（北京）科技有限公司 | A kind of detection of container lock and method for early warning based on deep learning |
| CN108074223A (en) * | 2017-12-28 | 2018-05-25 | 中国矿业大学(北京) | Fracture Networks extraction method in coal petrography sequence C T figures |
| CN109614913A (en) * | 2018-12-05 | 2019-04-12 | 北京纵目安驰智能科技有限公司 | A kind of oblique parking stall recognition methods, device and storage medium |
| CN111340763A (en) * | 2020-02-20 | 2020-06-26 | 浙江省交通规划设计研究院有限公司 | Method for rapidly measuring rock mass crushing degree of tunnel excavation face |
| CN113744219A (en) * | 2021-08-25 | 2021-12-03 | 绍兴文理学院 | Rock joint fracture overall complexity measurement and analysis method based on improved fractal theory |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473548A (en) * | 2013-09-22 | 2013-12-25 | 铁道第三勘察设计院集团有限公司 | Method for extracting fracture structure information by means of image processing and priori knowledge |
-
2014
- 2014-05-12 CN CN201410197540.3A patent/CN103942809A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473548A (en) * | 2013-09-22 | 2013-12-25 | 铁道第三勘察设计院集团有限公司 | Method for extracting fracture structure information by means of image processing and priori knowledge |
Non-Patent Citations (1)
| Title |
|---|
| 王卫星 等: "遥感图像中机场跑道的检查", 《重庆邮电大学学报（自然科学版）》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106023197A (en) * | 2016-05-18 | 2016-10-12 | 南京师范大学 | Automated identification and extraction method of vertical stratum |
| CN106023197B (en) * | 2016-05-18 | 2018-11-09 | 南京师范大学 | A kind of method of upright rock stratum automatic identification and extraction |
| CN106683060A (en) * | 2017-01-03 | 2017-05-17 | 北京大学 | Firing model removing microwave noise method switch |
| CN107680092A (en) * | 2017-10-12 | 2018-02-09 | 中科视拓（北京）科技有限公司 | A kind of detection of container lock and method for early warning based on deep learning |
| CN107680092B (en) * | 2017-10-12 | 2020-10-27 | 中科视拓（北京）科技有限公司 | Container lock catch detection and early warning method based on deep learning |
| CN108074223A (en) * | 2017-12-28 | 2018-05-25 | 中国矿业大学(北京) | Fracture Networks extraction method in coal petrography sequence C T figures |
| CN109614913A (en) * | 2018-12-05 | 2019-04-12 | 北京纵目安驰智能科技有限公司 | A kind of oblique parking stall recognition methods, device and storage medium |
| CN111340763A (en) * | 2020-02-20 | 2020-06-26 | 浙江省交通规划设计研究院有限公司 | Method for rapidly measuring rock mass crushing degree of tunnel excavation face |
| CN113744219A (en) * | 2021-08-25 | 2021-12-03 | 绍兴文理学院 | Rock joint fracture overall complexity measurement and analysis method based on improved fractal theory |
| CN113744219B (en) * | 2021-08-25 | 2024-04-05 | 绍兴文理学院 | Rock joint fracture overall complexity measurement analysis method based on improved fractal science |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103942809A (en) | Method for detecting joint fissures in rock images | |
| CN105427323B (en) | A kind of laser melting coating welding pool edge extraction method based on phase equalization | |
| US11443500B2 (en) | Image processing apparatus, image processing method, and program for detecting defect from image | |
| CN103745221A (en) | Two-dimensional code image correction method | |
| CN108399644A (en) | A kind of wall images recognition methods and its device | |
| CN109724988B (en) | PCB defect positioning method based on multi-template matching | |
| CN112857252B (en) | Tunnel image boundary line detection method based on reflectivity intensity | |
| CN104537342A (en) | Quick lane line detection method combined with ridge boundary detection and Hough transformation | |
| CN103345743A (en) | Image segmentation method for intelligent flaw detection of cell tail end | |
| CN106327464A (en) | Edge detection method | |
| CN102073872A (en) | Image-based method for identifying shape of parasite egg | |
| CN111127498A (en) | Canny edge detection method based on edge self-growth | |
| CN103914829A (en) | Method for detecting edge of noisy image | |
| CN109271882B (en) | Method for extracting color-distinguished handwritten Chinese characters | |
| CN104408721A (en) | Stamper image extracting method based on background density estimation | |
| JP2019087050A (en) | Structure maintenance management job support system | |
| CN102542259A (en) | Identification method for near-shore on-land water body | |
| CN102313740A (en) | Solar panel crack detection method | |
| CN105894501A (en) | Single-tree detection and crown describing method for high-resolution remote sensing image | |
| CN116309284A (en) | Slope top/bottom line extraction system and method | |
| CN108171771A (en) | The line drawing in a kind of combination external margin information and interior polymeric road draws generating algorithm | |
| CN104680506A (en) | Method and system for detecting boundary line along different directions | |
| CN111105408B (en) | Building surface crack detection method and system based on image processing | |
| CN115187744A (en) | Cabinet identification method based on laser point cloud | |
| CN110533682B (en) | Image edge real-time extraction method based on curvature filtering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140723 |
|
| RJ01 | Rejection of invention patent application after publication |