CN103870833A - Method for extracting and evaluating pavement crack based on concavity measurement - Google Patents
Method for extracting and evaluating pavement crack based on concavity measurement Download PDFInfo
- Publication number
- CN103870833A CN103870833A CN201410125302.1A CN201410125302A CN103870833A CN 103870833 A CN103870833 A CN 103870833A CN 201410125302 A CN201410125302 A CN 201410125302A CN 103870833 A CN103870833 A CN 103870833A
- Authority
- CN
- China
- Prior art keywords
- crack
- value
- concavity
- pixel
- point
- 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
Images
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for extracting and evaluating a pavement crack based on concavity measurement. The method mainly comprises the steps that concavity characteristic strength calculation and processing are carried out on a collected pavement crack image through the concavity measurement; a concavity characteristic strength value is enhanced by adopting the local contrast; preliminary noise reduction is carried out through area variance calculation, and statistics and connected domain noise reduction based on geometrical features are carried out on geometrical features of all connected domains of a characteristic strength image processed in a binarization mode; morphological expansion processing and frame extraction are carried out on a result image after geometrical feature noise reduction; statistics is carried out on the geometric sense of the crack in the image after frame extraction, pixel tracking is carried out on a crack target through a depth search algorithm, and a linear target characteristic value obtained through tracking is stored; image information of the crack is converted into graphic data vector information to obtain topological morphological characteristics of the crack, and finally analysis, statistics and description are carried out on the characteristics of the crack. The method can automatically and accurately extract the crack target and carry out analysis, statistics and description on the morphology of the crack.
Description
Technical field
The present invention relates to image processing field, relate in particular to a kind of pavement crack of estimating based on concavity and extract and assessment method.
Background technology
In recent years, along with the development of China Transportation Industry, particularly the construction of highway has obtained development faster, the mileage of highway and heavy goods vehicles quantity are all in quick increase, it is a problem becoming increasingly conspicuous that the road surface profile thereupon causing is deformed into, the analysis in road pavement crack is the important component part of evaluating Pavement Condition, carries out especially the basis of highway scientific maintenance.Crack is repeatedly to travel and roll after generation flow deformation, wearing and tearing, depression through doughnut in road surface, and the often groove of longitudinal band shape producing in driveway wheelpath, is road surface permanent deformation, is made up of the depression of wheelmark and the protuberance of both sides.At present, highway in China maintenance technique for investigation part has realized robotization, but many specific works must, by manually completing, not only be wanted barring traffic, also will drop into a large amount of man power and materials.Manual type efficiency is low, investment large, poor stability, affect traffic, and the precision of data is also difficult to be guaranteed, and along with the development of the hardware technology such as computing machine and image acquisition, automatic collection and the detection technique of pavement crack image are growing.
Pavement crack disease is the Common Diseases type of highway pavement, belongs to two-dimentional wire disease.Highway in China is mainly take bituminous pavement and cement concrete pavement as main, and wherein asphalt pavement crack Damage Types has: be full of cracks, block crack, longitudinal crack, transverse crack etc.Therefore automatic detection of pavement disease is a problem that is not easy solution, is also to improve the problem that highway maintenance detection automatization level must be faced simultaneously.
At present, automatic collection and the storage means of pavement crack image reach its maturity, and the pavement crack image of collection is subject to the impact of many factors, as (1) imaging gray scale inequality; (2) image blurring; (3) shadow problem and white line; (4) irregularity in non-crack, road surface; (5) random noise etc., makes the extraction and quite difficulty of identification in crack in image processing process.
At present, the extracting method in crack has: on pre-service basis, fracture image carries out Threshold segmentation, then extracts the linear feature on image, or adopts Texture classification, finally identifies crack target according to linear feature from former gray level image.
Separately having a kind of method is on figure image intensifying basis, adopts between maximum kind, inter-object distance criterion carries out Threshold segmentation to image, on segmentation result figure, extracts FRACTURE CHARACTERISTICS.
Also having a kind of method is on figure image intensifying basis, carries out image cut apart based on multiscale space model.
In said method, all adopt the dividing method fracture image based on threshold value to process, but thresholding method is very poor to the extraction effect in thinner in figure or distant crack, even miss part crack information, and in result, there is a large amount of noises, can make like this testing result undesirable, not reach the object that pavement crack detects automatically.
Summary of the invention
The technical problem to be solved in the present invention is for road pavement crack extract in prior art not accurate enough, perfect, and the defect that robustness is not strong provides a kind of assessment method that can automatically, accurately extract crack target and carry out fracture morphology analysis.
The technical solution adopted for the present invention to solve the technical problems is:
Provide a kind of pavement crack of estimating based on concavity to extract and assessment method, comprise the following steps:
S101, collection pavement crack image;
S102, to gather pavement crack image be normalized, make the gray-scale value of entire image between 0-255;
S103, extraction concavity characteristic strength: the window of selecting a certain size according to the image after normalized in step S102, the concavity characteristic strength value of the centre coordinate pixel of calculation window, in whole image-region from left to right, moving window from top to bottom, calculate the concavity characteristic measurement value of each coordinate pixel in whole figure, measure value is carried out to threshold process, be less than setting to 0 of threshold value, be more than or equal to the intensity level that is set to of threshold value, obtain concavity characteristic strength figure, and normalize to 0-255;
The enhancing processing of S104, concavity characteristic strength figure: be not 0 point to intensity level in concavity characteristic strength figure, calculate its local contrast, if current some local contrast value is less than setting contrast threshold, the intensity level of current point is set to 0, if be more than or equal to setting threshold, retain initial value; Crack concavity characteristic strength figure after treatment is enhanced;
S105, Local Deviation denoising: utilize Local Deviation method of discrimination to carry out denoising to strengthening pavement crack concavity characteristic strength figure after treatment;
S106, the pavement crack concavity characteristic strength figure after variance denoising is carried out to binary conversion treatment, judge whether the value of pixel is less than the threshold value of a setting, if, be set to 0, if not, be set to 255, obtained the concavity characteristic pattern after binary conversion treatment;
S107, the geometrical property of all connected domains in the concavity characteristic pattern after binary conversion treatment is added up, comprise the number of contained pixel in connected domain, length and the width of connected domain;
S108, the denoising of connected domain geometrical property: the threshold value of setting connected domain length, width and number of pixels, judge in figure whether each geometrical property value in single connected domain is less than the threshold value of setting, if be less than setting threshold, this connected domain pixel value is all set to 0; Otherwise, retain this connected domain; Finally obtain the result figure after geometrical property denoising;
S109, the result figure after geometrical property denoising is carried out to morphological dilations processing;
S110, the crack target after expansion process is carried out to skeletal extraction, become single pixel linear target with topological morphological character;
The geometric sense in crack in figure after S111, statistics skeletal extraction, utilize deep search algorithm, each single pixel linear target is carried out to pixel tracking, preserve and follow the tracks of the linear target eigenwert obtaining, comprise coordinate, branch node coordinate, branch's number of all pixels in each target, length and the width in linear crack, and the shared region area in block crack;
S112, the crack geometric sense information obtaining according to tracking, the image information in crack is converted to graph data Vector Message, obtain the topological morphological feature in crack, analyze, add up, describe and evaluate according to the topological morphological feature fracture geometrical property in this crack.
In method of the present invention, in step S103, selection window specifically comprises:
By in the image after normalized in step S102 a bit
as window Selection Center point, select respectively two sizeable wicket A(sizes to be
) and large window (size is
), the region in large window outside wicket is B;
In step S103, calculate current point
concavity characteristic strength value, comprise the following steps:
1. calculate the mean value of pixel in wicket A
;
2. calculate
and the difference in the B of region between other each point value:
;
If 3.
, concavity number variable
add 1; If
, a concavity number variable
remain unchanged;
4. in the B of region, total number of pixels is designated as
;
5. calculate the concavity number factor
;
6. to all satisfying condition in window
difference
be averaging concavity intensity level
If 7. each concavity characteristic measurement value of current point satisfies condition simultaneously
and
, current some intensity assignment is
, otherwise be set to 0, obtain current point after treatment
concavity characteristic strength value.
In method of the present invention, step S104 specifically comprises:
(1) set sizeable window, the point that concavity characteristic strength figure intermediate value is greater than 0 is aimed at the center of window, and in former gray-scale map, confines the image-region in this window respective coordinates region;
(2) in the concavity characteristic strength figure in window, statistical value is not 0 number of pixels
;
(3), in the former gray-scale map in window, be not 0 grey scale pixel value summation corresponding to coordinate to (2) step intermediate value
,
coordinate pixel corresponding in strength characteristic figure is not 0;
(4) the interior value of calculation window is not 0 zone leveling value
;
(5) number of pixels that in the concavity characteristic strength figure in window, statistical value is 0
;
(6) in the former gray-scale map in window, the grey scale pixel value summation that the coordinate that is 0 to (5) step intermediate value is corresponding
,
coordinate pixel corresponding in characteristic pattern is 0;
(7) the zone leveling value that in calculation window, value is 0
;
(8) calculate the ratio of two the zone leveling value difference values in current some place:
;
(9) judge ratio
if this ratio is less than a setting threshold, current point value is set to 0; If be more than or equal to setting threshold, retain current point value;
(10) window at whole figure from left to right, slides from top to bottom, and window center to aim at intensity map intermediate value be not 0 point at every turn, obtain contrast concavity characteristic strength after treatment figure, and be normalized to 0-255,
In method of the present invention, calculate the variance of crack concavity characteristic strength figure
;
Calculate the variance yields of the each connected domain of crack concavity characteristic strength figure
;
Calculate the poor of the variance yields of each connected domain and whole figure variance yields:
.If
, retain this connected domain; If
, remove whole connected domain.
In method of the present invention, in step S111, utilize deep search algorithm, each single pixel linear target carried out to pixel tracking and specifically comprise the following steps:
A) in whole figure, from left to right point by point search is not 0 pixel from top to bottom, judges whether it is end points, and the starting point using first end points finding as first tree crack is accessing points with this point of tense marker, sets number of pixels counter and adds 1;
B) centered by initial end points, travel through 8 neighborhoods of this pixel, mark is not wherein 0 pixel is accessing points, tree number of pixels counter adds 1;
C) continue traversal upper one 8 neighborhood territory pixels of gauge point, if in its 8 neighborhood point, be not 0 and unlabelled pixel only have one, set number of pixels counter and add 1, mark current pixel point; If value be not 0 and unlabelled pixel have 2 and more than, according to priority determine that point sets pixel impact point for next, rest of pixels point coordinate is stacked, and mark current pixel point, defining current point is node, node counts device adds 1;
D) get back to the c) step, continue search, until look for complete the whole pixels of tree-shaped crack trunk;
E) by the mode of first in first out, node is popped, the starting point using node as access, repeated execution of steps c), until accessed whole nodes;
F) the trunk pixel number of the crack tree that statistics has been accessed, branch node number and coordinate, form crack topology information, realized quantitative calculating and the statistics of pavement crack;
Continue to search in the drawings initial end points, after finding, get back to the b) step, the related data of adding up next tree-shaped crack, forms topology information; Until have access to all slits Pixel Information in view picture figure, follow the tracks of and finish.
The beneficial effect that the present invention produces is: fracture image of the present invention carries out single pixel and follows the tracks of detection, and the crack object pixel obtaining is corresponding with the center line in true crack, and crack extract and accurate positioning, be not subject to the interference of road sign, shade; By deep search algorithm, realize quantitative calculating and the statistics of pavement crack, for follow-up pavement patching provides quantification information and Data support; Crack of the present invention concavity characteristic measurement calculates simple, and computational complexity is low, only has plus and minus calculation, the characteristic strength obtaining is obvious, elimination and the denoising of pseudo-characteristic are simple and convenient, and the processing time is short, meet the time requirement of express highway pavement Crack Detection to computing; Adopt in addition the method that the present invention proposes to test a large amount of pavement images that gather, program is stable, and adaptivity is good, strong robustness, and all picture processings, without manual intervention, have been obtained satisfied pavement crack and have been detected and evaluating result.
Accompanying drawing explanation
Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing:
Fig. 1 is that the pavement crack estimated based on concavity of the embodiment of the present invention extracts and the process flow diagram of assessment method;
Fig. 2 is the pavement crack gray level image that the embodiment of the present invention gathers, and normalization sets to 0-255;
Fig. 3 is that embodiment of the present invention pavement crack concavity is estimated calculating schematic diagram;
Fig. 4 is that the embodiment of the present invention is carried out concavity to Fig. 2 and estimated the characteristic strength figure calculating;
Fig. 5 is that the local contrast of embodiment of the present invention concavity intensity level is calculated schematic diagram;
Fig. 6 carries out to Fig. 2 the result figure that contrast computing obtains for the embodiment of the present invention on Fig. 4 basis;
Fig. 7 is the result that the embodiment of the present invention adopts Local Deviation Fig. 6 to be carried out to denoising;
Fig. 8 is the binaryzation result figure of Fig. 7;
Fig. 9 utilizes connected domain geometrical property to carry out the result of denoising on Fig. 8 basis;
Figure 10 is the result of Fig. 9 being carried out to morphological dilations processing;
Figure 11 is the result of Figure 10 being carried out to skeletal extraction;
Figure 12 is that the geometric configuration of embodiment of the present invention fracture is described and statistics schematic diagram.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
The pavement crack that the embodiment of the present invention is estimated based on concavity extracts with assessment method and mainly comprises the following steps:
S101, collection pavement crack image;
S102, to gather pavement crack image be normalized, make the gray-scale value of entire image between 0-255, as shown in Figure 2;
S103: extract concavity characteristic strength: the window of selecting a certain size according to the image after normalized in step S102, the concavity characteristic strength value of the centre coordinate pixel of calculation window, in whole image-region from left to right, moving window from top to bottom, calculate the concavity characteristic measurement value (comprising concavity intensity level and the concavity number factor) of each coordinate pixel in whole figure, measure value is carried out to threshold process, be less than setting to 0 of threshold value, be more than or equal to the intensity level that is set to of threshold value.Obtain concavity characteristic strength figure, and normalize to 0-255;
In one embodiment of the present of invention, as shown in Figure 2, take a pixel wherein as the current point that calculates concavity intensity is as example.Using this point as window Selection Center point, select respectively two sizeable wicket A(sizes to be
) and large window (size is
), the region in large window outside wicket is B; As shown in Figure 3;
In one embodiment of the present of invention, to current point
, calculate its concavity intensity level, adopt following steps to calculate:
1. calculate the mean value of pixel in wicket A
.
2. calculate
and the difference in the B of region between other each point value:
;
If 3.
, concavity number variable
add 1; If
, a concavity number variable
remain unchanged;
4. in the B of region, total number of pixels is designated as
;
5. calculate the concavity number factor
;
6. to all satisfying condition in window
difference
be averaging concavity intensity level
If 7. each concavity characteristic measurement value of current point (comprising concavity intensity level and the concavity number factor) satisfies condition simultaneously
and
, current some intensity assignment is
, otherwise be set to 0, obtain current point after treatment
concavity characteristic strength value.
Moving window from top to bottom from left to right in whole image-region, obtains the concavity characteristic strength value of each coordinate pixel in whole figure;
After above-mentioned processing, obtain the concavity characteristic strength figure of former gray-scale map, as shown in Figure 4.
S104, meet step S103, on above-mentioned processing basis, concavity characteristic strength figure is carried out to region contrast and strengthen and process, step is as follows:
[1] set size to fit
window, the point that concavity characteristic strength figure intermediate value is greater than 0 is aimed at the center of window, and in former gray-scale map, confines the image-region in this window respective coordinates region;
[2] in the concavity characteristic strength figure in window, statistical value is not 0 number of pixels
;
[3], in the former gray-scale map in window, be not 0 grey scale pixel value summation corresponding to coordinate to (2) step intermediate value
,
coordinate pixel corresponding in strength characteristic figure is not 0;
[4] the interior intensity level of calculation window is not 0 zone leveling value
;
[5] number of pixels that in the concavity characteristic strength figure in window, statistical value is 0
;
[6] in the former gray-scale map in window, to the grey scale pixel value summation that in (5) step, intensity level is 0
,
the former figure gray-scale value that is 0 in intensity map intermediate value;
[7] the zone leveling value that in calculation window, value is 0
;
[8] calculate the ratio of two the zone leveling value difference values in current some place:
;
[9] judgement ratio above, if this ratio is less than a setting threshold, is set to 0 by current point value; If be more than or equal to setting threshold, retain current point value;
[10] window at whole figure from left to right, slides from top to bottom, and window center to aim at intensity map intermediate value be not 0 point at every turn, obtain contrast concavity characteristic strength after treatment figure, and be normalized to 0-255, result as shown in Figure 6.
S105, meet step S104, to above-mentioned crack concavity characteristic strength figure, adopt variance denoising method to remove pseudo-target;
(1) variance of calculating concavity intensity map (being crack concavity characteristic strength figure)
;
(2) variance yields of each connected domain in calculating concavity intensity map
;
(3) calculate the poor of the variance yields of each connected domain and whole figure variance yields:
.If
, retain this connected domain; If
, remove whole connected domain;
(4) obtain the result after Variance feature denoising, as shown in Figure 7;
S106, meet step S105, to the pavement crack concavity characteristic strength figure binaryzation after variance denoising, judge whether the value of pixel is 0, be if so, set to 0, if not, be set to 255, the concavity obtaining after binary conversion treatment is estimated figure, as shown in Figure 8;
S107, the geometrical property of each connected domain in view picture figure (the concavity intensity map after binary conversion treatment) is added up, comprise the number of contained pixel in each connected domain, the length of connected domain and width, and whole mean value of pixel grey scales in each connected domain;
S108, the denoising of connected domain geometrical property: set the threshold value of connected domain length, width and number of pixels, judge in figure in single connected domain that whether each eigenwert is less than the threshold value of setting, if be less than setting threshold, is set to 0 by this connected domain pixel value; Otherwise, retain this connected domain.As set connected domain length threshold
if, connected domain length in figure
be less than setting threshold, this connected domain pixel value be set to 0; Otherwise, retain this connected domain; Obtain the result images after geometrical property denoising, as shown in Figure 9;
S109, the result figure after geometrical property denoising is carried out to expansion process, the results are shown in Figure 10;
S110, the crack target after expansion process being carried out to skeletal extraction, become single pixel linear target with topological morphological character, as shown in figure 11, is skeletal extraction result;
The geometric sense in crack in figure after S111, statistics skeletal extraction.Utilize deep search algorithm, each crack is carried out to pixel tracking, preserve the coordinate of pixel in each target, branch node coordinate, branch's number, available schematic diagram Figure 12 describes.In one embodiment of the present of invention, concrete searching algorithm step is as follows:
A), in whole figure, from left to right point by point search is not 0 pixel from top to bottom, judges whether it is end points, starting point using first end points finding as first tree crack, as A point in Figure 12, be accessing points with this point of tense marker, tree number of pixels counter adds 1;
B) centered by initial end points, travel through 8 neighborhoods of this pixel, mark is not wherein 0 pixel is accessing points, tree number of pixels counter adds 1;
C) continue traversal upper one 8 neighborhood territory pixels of gauge point, if in its 8 neighborhood point, be not 0 and unlabelled pixel only have one, set number of pixels counter and add 1, mark current pixel point; If value be not 0 and unlabelled pixel have 2 and more than, according to priority determine that point sets pixel impact point for next, rest of pixels point coordinate is stacked, and mark current pixel point, defining current point is node, node counts device adds 1;
D) get back to the 3. step, continue search, until look for complete the whole pixels of tree-shaped crack trunk;
E) by the mode of first in first out, node is popped, the starting point using node as access, repeated execution of steps 3., until accessed whole nodes;
F) the trunk pixel number of the crack tree that statistics has been accessed, branch node number and coordinate, form topology information;
G) continue to search in the drawings initial end points, after finding, get back to the 2. step, the related data of adding up next tree-shaped crack, forms topology information; Until have access to all slits Pixel Information in view picture figure, follow the tracks of and finish.
S112, meet step S111, the linear target eigenwert obtaining according to tracking, the image information in crack is converted to graph data Vector Message, obtain the topological morphological feature in crack, fracture characteristic is analyzed, adds up, is described and evaluates, and data is submitted to highway administration department and carry out decision-making.
Compared with existing other pavement crack image detection algorithm, the method that the present invention proposes has the following advantages:
1. accuracy: fracture image of the present invention carries out single pixel to be followed the tracks of and detect, the crack object pixel obtaining is corresponding with the center line in true crack, crack extract and accurate positioning, and be not subject to the interference of road sign, shade;
2. quantification: the algorithm of chain code following is used in the analysis of processing fracture morphology geometric properties, has realized quantitative calculating and the statistics of pavement crack, for follow-up pavement patching provides quantification information and Data support;
3. rapidity and high efficiency: crack of the present invention concavity characteristic measurement calculates simple, and computational complexity is low, only has plus and minus calculation, and the characteristic strength obtaining is obvious, and elimination and the denoising of pseudo-characteristic are simple and convenient.On multi-purpose computer, every width image is no more than 0.5 second working time, uses special tool control computer to process, and the processing time is in 0.3 second, meets the time requirement of express highway pavement Crack Detection to computing;
4. adaptivity and robustness are good: adopt the method that the present invention proposes to test a large amount of pavement images that gather, program is stable, and adaptivity is good, strong robustness, all picture processings, without manual intervention, have been obtained satisfied pavement crack and have been detected effect.
Should be understood that, for those of ordinary skills, can be improved according to the above description or convert, and all these improvement and conversion all should belong to the protection domain of claims of the present invention.
Claims (5)
1. the pavement crack of estimating based on concavity extracts and an assessment method, it is characterized in that, comprises the following steps:
S101, collection pavement crack image;
S102, to gather pavement crack image be normalized, make the gray-scale value of entire image between 0-255;
S103, extraction concavity characteristic strength: the window of selecting a certain size according to the image after normalized in step S102, the concavity characteristic strength value of the centre coordinate pixel of calculation window, in whole image-region from left to right, moving window from top to bottom, calculate the concavity characteristic measurement value of each coordinate pixel in whole figure; Concavity characteristic measurement value is carried out to threshold process, be less than setting to 0 of threshold value, be more than or equal to the intensity level that is set to of threshold value, obtain concavity characteristic strength figure, and normalize to 0-255;
The enhancing processing of S104, concavity characteristic strength figure: be not 0 point to intensity level in concavity characteristic strength figure, calculate its local contrast, if current some local contrast value is less than setting contrast threshold, the intensity level of current point is set to 0, if be more than or equal to setting threshold, retain initial value; Crack concavity characteristic strength figure after treatment is enhanced;
S105, Local Deviation denoising: utilize Local Deviation method of discrimination to carry out denoising to strengthening pavement crack concavity characteristic strength figure after treatment;
S106, the pavement crack concavity characteristic strength figure after variance denoising is carried out to binary conversion treatment, judge whether the value of pixel is less than the threshold value of a setting, if, be set to 0, if not, be set to 255, obtained the concavity characteristic pattern after binary conversion treatment;
S107, the geometrical property of all connected domains in the concavity characteristic pattern after binary conversion treatment is added up, comprise the number of contained pixel in connected domain, length and the width of connected domain;
S108, the denoising of connected domain geometrical property: the threshold value of setting connected domain length, width and number of pixels, judge in figure whether each geometrical property value in single connected domain is less than the threshold value of setting, if be less than setting threshold, this connected domain pixel value is all set to 0; Otherwise, retain this connected domain; Finally obtain the result figure after geometrical property denoising;
S109, the result figure after geometrical property denoising is carried out to morphological dilations processing;
S110, the crack target after expansion process is carried out to skeletal extraction, become single pixel linear target with topological morphological character;
The geometric sense in crack in figure after S111, statistics skeletal extraction, utilize deep search algorithm, each single pixel linear target is carried out to pixel tracking, preserve and follow the tracks of the linear target eigenwert obtaining, comprise coordinate, branch node coordinate, branch's number of all pixels in each target, length and the width in linear crack, and the shared region area in block crack;
S112, the crack geometric sense information obtaining according to tracking, the image information in crack is converted to graph data Vector Message, obtain the topological morphological feature in crack, analyze, add up, describe and evaluate according to the topological morphological feature fracture geometrical property in this crack.
2. method according to claim 1, is characterized in that,
In step S103, selection window specifically comprises: by the image after normalized in step S102 a bit
as window Selection Center point, select respectively two sizeable wicket A and large window, wherein wicket size is
, large window size is
, the region in large window outside wicket is B;
In step S103, calculate current point
concavity characteristic strength value, comprise the following steps:
1. calculate the mean value of pixel in wicket A
;
2. calculate
and the difference in the B of region between other each point value:
;
If 3.
, concavity number variable
add 1; If
, a concavity number variable
remain unchanged;
4. in the B of region, total number of pixels is designated as
;
5. calculate the concavity number factor
;
6. to all satisfying condition in window
difference
be averaging concavity intensity level
If 7. each concavity characteristic measurement value of current point satisfies condition simultaneously
and
, current some intensity assignment is
, otherwise be set to 0, obtain current point after treatment
concavity characteristic strength value.
3. method according to claim 1, is characterized in that, step S104 specifically comprises:
Set sizeable window, the point that concavity characteristic strength figure intermediate value is greater than 0 is aimed at the center of window, and in former gray-scale map, confines the image-region in this window respective coordinates region;
In concavity characteristic strength figure in window, statistical value is not 0 number of pixels
;
In former gray-scale map in window, be not 0 grey scale pixel value summation corresponding to coordinate to (2) step intermediate value
,
coordinate pixel corresponding in strength characteristic figure is not 0;
In calculation window, value is not 0 zone leveling value
;
The number of pixels that in concavity characteristic strength figure in window, statistical value is 0
;
In former gray-scale map in window, the grey scale pixel value summation that the coordinate that is 0 to (5) step intermediate value is corresponding
,
coordinate pixel corresponding in characteristic pattern is 0;
The zone leveling value that in calculation window, value is 0
;
Calculate the ratio of two the zone leveling value difference values in current some place:
;
Judge ratio
if this ratio is less than a setting threshold, current point value is set to 0; If be more than or equal to setting threshold, retain current point value;
Window at whole figure from left to right, slides from top to bottom, and window center to aim at intensity map intermediate value be not 0 point at every turn, obtain contrast concavity characteristic strength after treatment figure, and be normalized to 0-255.
4. method according to claim 1, is characterized in that, step S105 specifically comprises the following steps:
Calculate the variance of crack concavity characteristic strength figure
;
Calculate the variance yields of the each connected domain of crack concavity characteristic strength figure
;
Calculate the poor of the variance yields of each connected domain and whole figure variance yields:
;
If
, retain this connected domain; If
, remove whole connected domain.
5. method according to claim 1, is characterized in that, utilizes deep search algorithm in step S111, each single pixel linear target is carried out to pixel tracking and specifically comprise the following steps:
In whole figure, from left to right point by point search is not 0 pixel from top to bottom, judges whether it is end points, and the starting point using first end points finding as first tree crack is accessing points with this point of tense marker, sets number of pixels counter and adds 1;
Centered by initial end points, travel through 8 neighborhoods of this pixel, mark is not wherein 0 pixel is accessing points, tree number of pixels counter adds 1;
Continue traversal upper one 8 neighborhood territory pixels of gauge point, if in its 8 neighborhood point, be not 0 and unlabelled pixel only have one, set number of pixels counter and add 1, mark current pixel point; If value be not 0 and unlabelled pixel have 2 and more than, according to priority determine that point sets pixel impact point for next, rest of pixels point coordinate is stacked, and mark current pixel point, defining current point is node, node counts device adds 1;
Get back to the c) step, continue search, until look for complete the whole pixels of tree-shaped crack trunk;
By the mode of first in first out, node is popped, the starting point using node as access, repeated execution of steps c), until accessed whole nodes;
The trunk pixel number of the crack tree that statistics has been accessed, branch node number and coordinate, form topology information;
Continue to search in the drawings initial end points, after finding, get back to the b) step, the related data of adding up next tree-shaped crack, forms topology information; Until have access to all slits Pixel Information in view picture figure, follow the tracks of and finish.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410125302.1A CN103870833B (en) | 2014-03-31 | 2014-03-31 | The pavement crack estimated based on concavity extracts and assessment method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410125302.1A CN103870833B (en) | 2014-03-31 | 2014-03-31 | The pavement crack estimated based on concavity extracts and assessment method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103870833A true CN103870833A (en) | 2014-06-18 |
| CN103870833B CN103870833B (en) | 2016-03-02 |
Family
ID=50909349
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410125302.1A Expired - Fee Related CN103870833B (en) | 2014-03-31 | 2014-03-31 | The pavement crack estimated based on concavity extracts and assessment method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103870833B (en) |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537652A (en) * | 2014-12-17 | 2015-04-22 | 中交第一公路勘察设计研究院有限公司 | Cement concrete pavement notch groove depth three-dimension detecting algorithm and system |
| CN104535589A (en) * | 2015-01-20 | 2015-04-22 | 广东电网有限责任公司电力科学研究院 | Online detection method and device for low-voltage current mutual inductor |
| CN104568981A (en) * | 2015-01-06 | 2015-04-29 | 中科华核电技术研究院有限公司北京分公司 | Building wall defect measurement method based on image and system thereof |
| CN104657706A (en) * | 2015-01-27 | 2015-05-27 | 北京大学 | Image-based high-speed railway line pole breakage abnormality and connecting structural body abnormality detection method |
| CN104992429A (en) * | 2015-04-23 | 2015-10-21 | 北京宇航时代科技发展有限公司 | Mountain crack detection method based on image local reinforcement |
| CN105842239A (en) * | 2016-03-23 | 2016-08-10 | 北京盛广拓公路科技有限公司 | Asphalt aging performance evaluation method |
| CN105913689A (en) * | 2016-06-28 | 2016-08-31 | 成都之达科技有限公司 | Pre-warning method for motor vehicle driving safety |
| WO2016172827A1 (en) * | 2015-04-27 | 2016-11-03 | 武汉武大卓越科技有限责任公司 | Stepwise-refinement pavement crack detection method |
| CN107886542A (en) * | 2017-12-10 | 2018-04-06 | 上海交通大学 | The target restoration methods and system of a kind of vision-based detection |
| CN108074256A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method, apparatus and system based on distribution process |
| CN108074258A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method, apparatus and system based on parallel processing |
| CN108074225A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method and device |
| CN108074254A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method, apparatus and system based on stream handle |
| JPWO2017110279A1 (en) * | 2015-12-25 | 2018-10-04 | 富士フイルム株式会社 | Damage information processing apparatus and damage information processing method |
| CN109522646A (en) * | 2018-11-15 | 2019-03-26 | 中国矿业大学 | Rock mass intersection crack image is automatically separated and its vector quantization expression |
| CN109839439A (en) * | 2019-03-21 | 2019-06-04 | 陈朝忠 | Road and bridge pavement crack detection system and its detection method |
| CN110390256A (en) * | 2019-05-29 | 2019-10-29 | 广东建科交通工程质量检测中心有限公司 | A kind of asphalt pavement crack extracting method |
| CN110473187A (en) * | 2019-08-08 | 2019-11-19 | 武汉武大卓越科技有限责任公司 | A kind of line scanning three-dimensional pavement crack extract method of object-oriented |
| CN110575176A (en) * | 2019-08-26 | 2019-12-17 | 南京理工大学 | Action segmentation method based on two-layer sliding window threshold |
| CN112598638A (en) * | 2020-12-22 | 2021-04-02 | 东南大学 | Pavement crack detection method based on laser point cloud data |
| CN113486896A (en) * | 2021-07-22 | 2021-10-08 | 成都圭目机器人有限公司 | Automatic connection method for intermittent cracks of pavement image |
| CN114486939A (en) * | 2022-04-08 | 2022-05-13 | 欧普康视科技股份有限公司 | Lens scratch detection system and method |
| CN115409825A (en) * | 2022-09-06 | 2022-11-29 | 重庆众仁科技有限公司 | Temperature-humidity pressure trace identification method based on image identification |
| CN117152187A (en) * | 2023-10-30 | 2023-12-01 | 山东中科冶金矿山机械有限公司 | Crack contour extraction method in geological mapping |
| CN117522195A (en) * | 2023-11-06 | 2024-02-06 | 北京住总集团有限责任公司 | Residential quality detection and assessment system and method |
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 |
| CN101788288A (en) * | 2010-02-11 | 2010-07-28 | 长安大学 | Positioning system and positioning method for pavement cracks |
-
2014
- 2014-03-31 CN CN201410125302.1A patent/CN103870833B/en not_active Expired - Fee Related
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 |
| CN101788288A (en) * | 2010-02-11 | 2010-07-28 | 长安大学 | Positioning system and positioning method for pavement cracks |
Non-Patent Citations (1)
| Title |
|---|
| 刘凡凡等: "基于连通域相关及Hough变换的公路路面裂缝提取", 《北京邮电大学学报》, vol. 32, no. 2, 30 April 2009 (2009-04-30), pages 24 - 28 * |
Cited By (43)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104537652B (en) * | 2014-12-17 | 2017-06-20 | 中交第一公路勘察设计研究院有限公司 | A kind of cement-concrete road-surface groove-carving depth 3 D detection method and system |
| CN104537652A (en) * | 2014-12-17 | 2015-04-22 | 中交第一公路勘察设计研究院有限公司 | Cement concrete pavement notch groove depth three-dimension detecting algorithm and system |
| CN104568981A (en) * | 2015-01-06 | 2015-04-29 | 中科华核电技术研究院有限公司北京分公司 | Building wall defect measurement method based on image and system thereof |
| CN104535589A (en) * | 2015-01-20 | 2015-04-22 | 广东电网有限责任公司电力科学研究院 | Online detection method and device for low-voltage current mutual inductor |
| CN104657706A (en) * | 2015-01-27 | 2015-05-27 | 北京大学 | Image-based high-speed railway line pole breakage abnormality and connecting structural body abnormality detection method |
| CN104657706B (en) * | 2015-01-27 | 2017-12-29 | 北京大学 | The fracture of high ferro circuit line bar and draw bail body method for detecting abnormality based on image |
| CN104992429A (en) * | 2015-04-23 | 2015-10-21 | 北京宇航时代科技发展有限公司 | Mountain crack detection method based on image local reinforcement |
| CN104992429B (en) * | 2015-04-23 | 2017-09-12 | 北京宇航时代科技发展有限公司 | One kind is based on the enhanced mountain cracks detection method of image local |
| AU2015392660B2 (en) * | 2015-04-27 | 2019-05-16 | Wuhan Optics Valley Zoyon Science And Technology Co., Ltd. | Stepwise-refinement pavement crack detection method |
| WO2016172827A1 (en) * | 2015-04-27 | 2016-11-03 | 武汉武大卓越科技有限责任公司 | Stepwise-refinement pavement crack detection method |
| US10489903B2 (en) | 2015-04-27 | 2019-11-26 | Wuhan Wuda Zoyon Science And Technology Co., Ltd. | Stepwise refinement detection method for pavement cracks |
| JPWO2017110279A1 (en) * | 2015-12-25 | 2018-10-04 | 富士フイルム株式会社 | Damage information processing apparatus and damage information processing method |
| CN105842239B (en) * | 2016-03-23 | 2018-08-14 | 北京盛广拓公路科技有限公司 | A kind of Evaluation method of asphalt aging properties |
| CN105842239A (en) * | 2016-03-23 | 2016-08-10 | 北京盛广拓公路科技有限公司 | Asphalt aging performance evaluation method |
| CN105913689A (en) * | 2016-06-28 | 2016-08-31 | 成都之达科技有限公司 | Pre-warning method for motor vehicle driving safety |
| CN105913689B (en) * | 2016-06-28 | 2018-08-28 | 成都之达科技有限公司 | Motor vehicle running safe early warning method |
| CN108074225A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method and device |
| CN108074225B (en) * | 2016-11-11 | 2022-03-08 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extraction method and device |
| CN108074258B (en) * | 2016-11-11 | 2022-03-08 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extraction method, device and system based on parallel processing |
| CN108074258A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method, apparatus and system based on parallel processing |
| CN108074254B (en) * | 2016-11-11 | 2022-03-04 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extraction method, device and system based on stream processor |
| CN108074256B (en) * | 2016-11-11 | 2022-03-04 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extraction method, device and system based on distributed processing |
| CN108074256A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method, apparatus and system based on distribution process |
| CN108074254A (en) * | 2016-11-11 | 2018-05-25 | 中国石油化工股份有限公司抚顺石油化工研究院 | Sulfide information extracting method, apparatus and system based on stream handle |
| CN107886542A (en) * | 2017-12-10 | 2018-04-06 | 上海交通大学 | The target restoration methods and system of a kind of vision-based detection |
| CN109522646A (en) * | 2018-11-15 | 2019-03-26 | 中国矿业大学 | Rock mass intersection crack image is automatically separated and its vector quantization expression |
| CN109839439A (en) * | 2019-03-21 | 2019-06-04 | 陈朝忠 | Road and bridge pavement crack detection system and its detection method |
| CN109839439B (en) * | 2019-03-21 | 2021-08-20 | 广西安盛建设工程检测咨询有限公司 | Road and bridge pavement crack detection system and detection method thereof |
| CN110390256B (en) * | 2019-05-29 | 2023-06-23 | 广东建科交通工程质量检测中心有限公司 | Asphalt pavement crack extraction method |
| CN110390256A (en) * | 2019-05-29 | 2019-10-29 | 广东建科交通工程质量检测中心有限公司 | A kind of asphalt pavement crack extracting method |
| CN110473187A (en) * | 2019-08-08 | 2019-11-19 | 武汉武大卓越科技有限责任公司 | A kind of line scanning three-dimensional pavement crack extract method of object-oriented |
| CN110575176A (en) * | 2019-08-26 | 2019-12-17 | 南京理工大学 | Action segmentation method based on two-layer sliding window threshold |
| CN110575176B (en) * | 2019-08-26 | 2024-03-26 | 南京理工大学 | Action segmentation method based on two-layer sliding window threshold |
| CN112598638A (en) * | 2020-12-22 | 2021-04-02 | 东南大学 | Pavement crack detection method based on laser point cloud data |
| CN112598638B (en) * | 2020-12-22 | 2024-04-09 | 东南大学 | Pavement crack detection method based on laser point cloud data |
| CN113486896A (en) * | 2021-07-22 | 2021-10-08 | 成都圭目机器人有限公司 | Automatic connection method for intermittent cracks of pavement image |
| CN113486896B (en) * | 2021-07-22 | 2023-06-20 | 成都圭目机器人有限公司 | Automatic connection method for intermittent cracks of road surface image |
| CN114486939A (en) * | 2022-04-08 | 2022-05-13 | 欧普康视科技股份有限公司 | Lens scratch detection system and method |
| CN115409825A (en) * | 2022-09-06 | 2022-11-29 | 重庆众仁科技有限公司 | Temperature-humidity pressure trace identification method based on image identification |
| CN115409825B (en) * | 2022-09-06 | 2023-09-12 | 重庆众仁科技有限公司 | Temperature, humidity and pressure trace identification method based on image identification |
| CN117152187A (en) * | 2023-10-30 | 2023-12-01 | 山东中科冶金矿山机械有限公司 | Crack contour extraction method in geological mapping |
| CN117152187B (en) * | 2023-10-30 | 2024-01-26 | 山东中科冶金矿山机械有限公司 | Crack contour extraction method in geological mapping |
| CN117522195A (en) * | 2023-11-06 | 2024-02-06 | 北京住总集团有限责任公司 | Residential quality detection and assessment system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103870833B (en) | 2016-03-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103870833B (en) | The pavement crack estimated based on concavity extracts and assessment method | |
| Zhang et al. | Automatic pavement defect detection using 3D laser profiling technology | |
| CN112419250B (en) | Pavement crack digital image extraction, crack repair and crack parameter calculation method | |
| CN107092877B (en) | Remote sensing image roof contour extraction method based on building base vector | |
| AU2015392660B2 (en) | Stepwise-refinement pavement crack detection method | |
| CN105787937B (en) | A kind of high resolution remote sensing image road change detecting method based on OSM | |
| CN104008553A (en) | Crack detection method with image gradient direction histogram and watershed method conflated | |
| CN114998852A (en) | Intelligent detection method for road pavement diseases based on deep learning | |
| CN108921076B (en) | Pavement crack disease self-adaptive constant false alarm detection method based on image | |
| CN101551851A (en) | Infrared image target identification method | |
| CN104182754A (en) | Rural resident point information extraction method based on high-resolution remote-sensing image | |
| CN108898101B (en) | High-resolution SAR image road network detection method based on sketch and prior constraint | |
| CN103048329A (en) | Pavement crack detecting method based on active contour model | |
| CN104183142A (en) | Traffic flow statistics method based on image visual processing technology | |
| CN114463286A (en) | Road surface point cloud smoothing and crack information processing method based on moving window | |
| CN104680538A (en) | SAR image CFAR target detection method on basis of super pixels | |
| CN104992429A (en) | Mountain crack detection method based on image local reinforcement | |
| CN104899892A (en) | Method for quickly extracting star points from star images | |
| Chen et al. | An improved minimal path selection approach with new strategies for pavement crack segmentation | |
| CN112085699B (en) | Pavement crack extraction method based on two-dimensional image | |
| CN109540925A (en) | Complicated ceramic tile surface defect inspection method based on difference shadow method and local variance measurement operator | |
| CN109870458B (en) | Pavement crack detection and classification method based on three-dimensional laser sensor and bounding box | |
| CN113378659A (en) | Method for identifying busy operation | |
| CN109815961A (en) | A kind of pavement patching class Defect inspection method based on local grain binary pattern | |
| CN113240735A (en) | Slope displacement activity monitoring method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160302 Termination date: 20170331 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |