CN107578414A - A kind of processing method of pavement crack image - Google Patents
A kind of processing method of pavement crack image Download PDFInfo
- Publication number
- CN107578414A CN107578414A CN201710710006.1A CN201710710006A CN107578414A CN 107578414 A CN107578414 A CN 107578414A CN 201710710006 A CN201710710006 A CN 201710710006A CN 107578414 A CN107578414 A CN 107578414A
- Authority
- CN
- China
- Prior art keywords
- pavement crack
- image
- crack image
- region
- iteration
- 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
Landscapes
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of processing method of pavement crack image, mainly comprising herein below：1. by iterative process, pavement crack image is handled using class variance method between maximum, and the untrusted region in each iteration result is eliminated in pavement crack image；2. less iterations is used as the Preprocessing Algorithm of pavement crack image；3. road pavement crack pattern picture carries out excessive iteration,, as drawing of seeds picture, the discreet region in drawing of seeds picture will be eliminated, and utilize edge detection algorithm after result treatment, " growth fence " is set in pavement crack image, image segmentation is completed finally by region seed growth.Pavement crack image processing method proposed by the present invention significantly improves image preprocessing effect and image segmentation, overcome the interference of noisy noise in image, computational efficiency is higher, and there is stronger robustness, stable performance, applied to the Intelligent Recognition and statistics of pavement crack disease, the accuracy of identification and statistical efficiency of pavement crack disease are improved.
Description
Technical field
The invention belongs to field of road and technical field of image processing, and in particular to a kind of place of pavement crack image
Reason method.
Background technology
By the end of the year 2016, national highway total kilometrage ranks first in the world up to 130,000 kilometers.The extensive of highway is built
It is located at and daily life is also contributed to while affecting economic development, this causes the detection of highway and maintenance management also day
Benefit highlights importance and urgency.At present, with sensor, automatically control, the development of the technology such as computer, pavement image
Convergence is ripe for automatic acquisition equipment, and by the way of the crack identification in later stage is artificial still using man-computer cooperation or even completely, work
Work amount is big, efficiency is low.The pavement crack image collected for detection car develops algorithm, and place is identified in road pavement crack pattern picture
Reason can greatly improve the detection efficiency of pavement crack, and save human resources.
Pavement crack recognition methods common at present mainly has following several:
(1) the crack identification method based on gray scale threshold value, analyzed by road pavement gradation of image feature, it is suitable to choose
Gray scale threshold value distinguish image background and target.This method is typically found at generally lower than the background gray scale premise of gray scale in crack
Under the conditions of, it is desired to crack has higher contrast and a preferable continuity, but due to road surface dust stratification, crack slotted wall come off, road
Reason, the cracks such as face grain texture is abundant generally have the features such as low contrast, poor continuity, therefore based on the crack of gray scale threshold value
Recognition methods is difficult to not significant enough the disease of gray feature.
(2) the crack identification method based on Morphological scale-space, this method utilize burn into expansion, skeletal extraction, rim detection
The methods of obtain crack Two-dimensional morphology feature.But pavement image is complicated, disease form is various, the knowledge based on Morphological scale-space
Other method practicality is not high.
(3) the crack identification method based on machine learning, this method are mainly used in the classification of type after Crack Detection, crucial
It is the extraction of pavement crack feature and the design of grader.Because road conditions are complicated, crack form is various, FRACTURE CHARACTERISTICS extraction is difficult
Degree increases, while test sample collection is smaller, algorithm is complicated, the factor such as computationally intensive all governs the accuracy of sorting algorithm, Shandong
Rod and real-time.
(4) recognition methods of the pavement crack based on multi-scale geometric analysis, image geometry architectural feature generally is utilized, is adopted
With small echo, Ridge1et (ridge ripple), Curve1et (curve ripple), Contour1et (profile ripple), Bande1et (tape ripple) etc.
Map table reaches image information.Because the asphalt pavement crack under complex background has scrambling, fracture pattern and position have
Unpredictability, this method can not effectively extract complex fracture information, meanwhile, multiscale analysis method generally existing calculated
Journey is complicated, it is less efficient the problems such as.
These existing methods, universal road pavement crack image quality requirements are higher, but pavement crack figure in Practical Project
As being not (day night) obtained under identical illumination condition, (fine day/cloudy), interference is contained in some of images
The part of graphical analysis, such as stochastic particle shape texture, non-uniform lighting and irregular road surface shade, noisy environment line
Bar, water mark, tire trace, oil stain etc..Extraction of these disturbing factors for crack can produce obvious influence.
The content of the invention
For the present invention to eliminate the lot of interfering factors in pavement crack image, more preferable road pavement crack pattern picture carries out image
Processing, pavement crack image of the present invention acquired in road detection vehicle CCD camera are provided at a kind of pavement crack image
Reason method, the interference of complex background in pavement crack image can be overcome.
In order to solve the above-mentioned technical problem, the invention provides a kind of processing method of pavement crack image, including it is following
Step：
Step 1, using pavement crack and similitude of the road surface on gray value, pass through iterative process, control iterations
And/or iteration limit value, road pavement crack pattern picture carry out image segmentation pretreatment；
Step 2, obtain drawing of seeds picture, there is provided on the pavement crack image of " growth fence " carry out region seed life
It is long；
Wherein, the step 1 includes：By iterative process, handled using OTSU algorithm road pavement crack pattern pictures, and
Untrusted region in each iteration result is eliminated in pavement crack image；
Iterative process described in step 1 comprises the following steps：
Step 1.1：Using the gray value of each pixels of pavement crack image P after gray processing, pavement crack image is calculated
Gray average；
Step 1.2：Class variance method road pavement crack pattern picture carries out image segmentation between maximum based on image grey level histogram,
High luminance value regions are divided out, and road surface and pavement crack are divided into background area, wherein, the high luminance value regions refer to
Logical values 1 include roadmarking and frame, and the background area refers to logical values 0；
Step 1.3：8 connected regions for being divided into logical values 1 in image segmentation result, in pavement crack
Its gray value is uniformly set to the gray average being calculated in step 1.1 in image P；
Repeat step 1.1~1.3, to be repeated once step 1.1~1.3 as an iteration process.
Further, excessive iteration is carried out by road pavement crack pattern picture, using the result after iteration as drawing of seeds picture.
Further, " growth fence " is set in pavement crack image by edge detection algorithm.
Further, step 2 specifically includes following steps：
Step 3.1：Pavement crack image P after gray processing is iterated according to the iterative process in step 1, control changes
Generation number, binary conversion treatment is carried out using OTSU methods, obtains the drawing of seeds picture for region seed growth；
Step 3.2：Remove area in obtained drawing of seeds picture and be more than the connected region for limiting pixel, by the kind after processing
Subgraph is as the seed used in the seed growth of region；
Step 3.3：Using Canny rim detections for the pavement crack image P processing after gray processing, for gained
Drawing of seeds picture, the position pixel value on the corresponding pavement crack image P after gray processing in its region of logical values 1 is set to
It is capable of the value of restricted area seed growth；
Step 3.4：Region seed growth computing is done in the result images of step 3.3 with the seed in step 3.2, is obtained
To image segmentation result.
Further, the iterations in step 3.1 is higher than the iterations in step 1.
The beneficial effects of the invention are as follows：The present invention compared with prior art, has advantages below：The present invention is big by observing
Pavement strip region is measured, summing up graticule has the features such as gray value is larger, texture is shallower, shape is relatively regular, according to the above
Feature, the iteration OSTU algorithms employed in the present invention can effectively eliminate roadmarking in pavement crack image, overcome
The interference of roadmarking.
When iteration OSTU methods provided by the invention are used for the image preprocessing of pavement crack image, algorithm can be effective
The adverse effect that the noisy background of pavement crack image is split for image is eliminated, arithmetic speed is fast, and has very high robustness.
Brief description of the drawings
Fig. 1 is the algorithm flow explanation figure of the present invention；
Fig. 2 is the input picture gray level pavement crack image P in the specific embodiment of the invention；
Fig. 3 is the pretreated pavement crack image in the specific embodiment of the invention；
Fig. 4 is drawing of seeds picture of the removing in the specific embodiment of the invention less than 10 pixel connected regions；
Fig. 5 is the pavement crack image split by image in the specific embodiment of the invention.
Embodiment
The present embodiment be based on it is assumed hereinafter that, come realize the image of road pavement crack pattern picture split：
1st, crack pixel is deeper compared with road pixel；
2nd, the intensity profile of crack on road and road surface is independent；
3rd, a crack is a narrow, continuous target object；
4th, a crack is the section that one group interconnected, direction is different；
5th, the width of a crack is not steady state value over the entire length；
Based on assumed above, pavement crack and similitude of the road surface on gray value are utilized.The present embodiment uses iteration
Mode road pavement crack pattern picture carries out progressive processing, obtains drawing of seeds picture, rejects after noise there is provided the road of " growth fence "
Region seed growth is carried out on facial cleft seam image, obtains segmentation effect, intensity histogram of the present embodiment independent of gray level image
Figure, also has good adaptability for low contrast and uneven illumination situation.
Referring to Fig. 1 to Fig. 5, the present embodiment specifically includes following steps：
Step 1, the progressive processing of mode road pavement crack pattern picture progress using iteration：
Step 1.1：Using the gray value of each pixels of pavement crack image P after gray processing, pavement crack image is calculated
Gray average；
Step 1.2：Class variance method road pavement crack pattern picture carries out image segmentation between maximum based on image grey level histogram
Because pavement crack and road surface have similitude on gray value, therefore the high luminance value regions such as roadmarking are divided out
(roadmarking is set to logical values 1 with frame), road surface and pavement crack are divided into background area and (are set to logical
Value 0)；
Step 1.3：8 connected regions for being divided into logical values 1 in image segmentation result, it is considered to be non-
Putting property region, its gray value is uniformly set to the gray average being calculated in step 1.1 in pavement crack image P, with up to
To the effect for eliminating untrusted region；
Repeat step 1.1~1.3, to be repeated once step 1.1~1.3 as an iteration process, it is iterated with this.
Step 2, image is pre-processed：
By controlling limit value or iterations, when limit value is larger, when iterations is less, can effectively uniform
Noisy background in pavement crack image, further image segmentation is convenient for, is that a kind of effective pavement crack image is pre-
Processing method.
It is 3 times that iterations is controlled in the present embodiment, obtains the pavement crack image by pretreatment.
Step 3, there is provided on the pavement crack image of " growth fence " carry out region seed growth, obtain segmentation effect
Fruit：
Step 3.1：Excessive iterative processing is carried out for image using iteration OTSU methods, iteration time is controlled in the present embodiment
Number is 30 times, reuses OTSU methods and carries out binary conversion treatment, obtains result and is used for region seed growth, is referred to as " drawing of seeds
Picture "；
Step 3.2：Remove area in obtained drawing of seeds picture and be more than 8 connected regions for limiting pixel, after processing
Drawing of seeds picture limits pixel as 10 pixels as the seed used in the seed growth of region in the present embodiment；Because road surface
There is many noise regions in crack pattern picture, the production of the noise in subsequent step can effectively be suppressed by removing the seed of small area
It is raw；
Step 3.3：Using Canny rim detections for the pavement crack image P processing after gray processing, for gained
Drawing of seeds picture, the position pixel value on the corresponding pavement crack image P after gray processing in its region of logical values 1 is set to
The gray value upper limit 255, referred to as grows fence, this step can in region growing control area size, the mistake of inhibition zone
Degree growth, reach the effect for further suppressing noise；
Step 3.4：Region seed growth computing is done with the seed in step 3.2 in the result images of step 3.3 (to set
Gray scale difference is put 3), to obtain image segmentation result.
Claims (5)
- A kind of 1. processing method of pavement crack image, it is characterised in that：Comprise the following steps：Step 1, using pavement crack and similitude of the road surface on gray value, by iterative process, control iterations and/or Iteration limit value, road pavement crack pattern picture carry out image segmentation pretreatment；Step 2, obtain drawing of seeds picture, there is provided on the pavement crack image of " growth fence " carry out region seed growth；Wherein, the step 1 includes：By iterative process, handled using maximum variance between clusters road pavement crack pattern picture, And the untrusted region in each iteration result is eliminated in pavement crack image；Wherein, the iterative process described in step 1 comprises the following steps：Step 1.1：Using the gray value of each pixels of pavement crack image P after gray processing, the ash of calculating pavement crack image Spend average；Step 1.2：Class variance method road pavement crack pattern picture carries out image segmentation between maximum based on image grey level histogram, highlights Angle value region is divided out, and road surface and pavement crack are divided into background area, wherein, the high luminance value regions refer to Logical values 1 include roadmarking and frame, and the background area refers to logical values 0；Step 1.3：8 connected regions for being divided into logical values 1 in image segmentation result, in pavement crack image P It is middle that its gray value is uniformly set to the gray average being calculated in step 1.1；Repeat step 1.1~1.3, to be repeated once step 1.1~1.3 as an iteration process.
- A kind of 2. processing method of pavement crack image according to claim 1, it is characterised in that：Pass through road pavement crack Image carries out excessive iteration, using the result after iteration as drawing of seeds picture.
- A kind of 3. processing method of pavement crack image according to claim 1 or 2, it is characterised in that：Examined by edge Method of determining and calculating sets " growth fence " in pavement crack image.
- A kind of 4. processing method of pavement crack image according to claim 1, it is characterised in that：The step 2 is specific Comprise the following steps：Step 3.1：Pavement crack image P after gray processing is iterated according to the iterative process in step 1, control iteration time Number, reuse OTSU methods and carry out binary conversion treatment, obtain the drawing of seeds picture for region seed growth；Step 3.2：Remove area in obtained drawing of seeds picture and be more than the connected region for limiting pixel, by the drawing of seeds after processing As the seed used in the seed growth of region；Step 3.3：Using Canny rim detections for the pavement crack image P processing after gray processing, for the kind of gained Subgraph, the position pixel value on the corresponding pavement crack image P after gray processing in its region of logical values 1 is set to can The value of restricted area seed growth, referred to as grows fence；Step 3.4：Region seed growth computing is done in the result images of step 3.3 with the seed in step 3.2, obtains image Segmentation result.
- A kind of 5. processing method of pavement crack image according to right wants 4, it is characterised in that：In the step 3.1 Iterations is higher than the iterations in step 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710710006.1A CN107578414B (en) | 2017-08-18 | 2017-08-18 | Method for processing pavement crack image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710710006.1A CN107578414B (en) | 2017-08-18 | 2017-08-18 | Method for processing pavement crack image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107578414A true CN107578414A (en) | 2018-01-12 |
| CN107578414B CN107578414B (en) | 2021-09-07 |
Family
ID=61033804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710710006.1A Active CN107578414B (en) | 2017-08-18 | 2017-08-18 | Method for processing pavement crack image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107578414B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109632822A (en) * | 2018-12-25 | 2019-04-16 | 东南大学 | A kind of quasi-static high-precision road surface breakage intelligent identification device and its method |
| CN110555389A (en) * | 2019-08-09 | 2019-12-10 | 南京工业大学 | bullet line-bore trace identification method based on ridgelet transformation and rotation matching |
| CN110956183A (en) * | 2019-11-04 | 2020-04-03 | 东南大学 | Asphalt pavement crack morphology extraction method |
| CN118071752A (en) * | 2024-04-24 | 2024-05-24 | 中铁电气化局集团有限公司 | Contact net detection method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1793350A1 (en) * | 2005-12-01 | 2007-06-06 | Medison Co., Ltd. | Ultrasound imaging system and method for forming a 3D ultrasound image of a target object |
| CN104021574A (en) * | 2014-07-04 | 2014-09-03 | 武汉武大卓越科技有限责任公司 | Method for automatically identifying pavement diseases |
| CN104637047A (en) * | 2013-11-13 | 2015-05-20 | 北京慧眼智行科技有限公司 | Image processing method and device |
| CN106295789A (en) * | 2015-06-10 | 2017-01-04 | 浙江托普云农科技股份有限公司 | A kind of crop seed method of counting based on image procossing |
| CN106651872A (en) * | 2016-11-23 | 2017-05-10 | 北京理工大学 | Prewitt operator-based pavement crack recognition method and system |
-
2017
- 2017-08-18 CN CN201710710006.1A patent/CN107578414B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1793350A1 (en) * | 2005-12-01 | 2007-06-06 | Medison Co., Ltd. | Ultrasound imaging system and method for forming a 3D ultrasound image of a target object |
| CN104637047A (en) * | 2013-11-13 | 2015-05-20 | 北京慧眼智行科技有限公司 | Image processing method and device |
| CN104021574A (en) * | 2014-07-04 | 2014-09-03 | 武汉武大卓越科技有限责任公司 | Method for automatically identifying pavement diseases |
| CN106295789A (en) * | 2015-06-10 | 2017-01-04 | 浙江托普云农科技股份有限公司 | A kind of crop seed method of counting based on image procossing |
| CN106651872A (en) * | 2016-11-23 | 2017-05-10 | 北京理工大学 | Prewitt operator-based pavement crack recognition method and system |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109632822A (en) * | 2018-12-25 | 2019-04-16 | 东南大学 | A kind of quasi-static high-precision road surface breakage intelligent identification device and its method |
| CN110555389A (en) * | 2019-08-09 | 2019-12-10 | 南京工业大学 | bullet line-bore trace identification method based on ridgelet transformation and rotation matching |
| CN110555389B (en) * | 2019-08-09 | 2022-02-25 | 南京工业大学 | Bullet line-bore trace identification method based on ridgelet transformation and rotation matching |
| CN110956183A (en) * | 2019-11-04 | 2020-04-03 | 东南大学 | Asphalt pavement crack morphology extraction method |
| CN110956183B (en) * | 2019-11-04 | 2023-05-02 | 东南大学 | Asphalt pavement crack morphology extraction method |
| CN118071752A (en) * | 2024-04-24 | 2024-05-24 | 中铁电气化局集团有限公司 | Contact net detection method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107578414B (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2016172827A1 (en) | Stepwise-refinement pavement crack detection method | |
| CN107230202B (en) | Automatic identification method and system for road surface disease image | |
| CN104792792B (en) | A kind of road surface crack detection method of Stepwise Refinement | |
| CN113592861B (en) | Bridge crack detection method based on dynamic threshold | |
| CN107578414A (en) | A kind of processing method of pavement crack image | |
| CN104021574A (en) | Method for automatically identifying pavement diseases | |
| CN110956183B (en) | Asphalt pavement crack morphology extraction method | |
| CN108711149B (en) | Mineral rock granularity detection method based on image processing | |
| CN109064411B (en) | Illumination compensation-based road surface image shadow removing method | |
| CN109410205B (en) | Crack extraction method under complex pavement background | |
| CN106447673A (en) | Chip pin extraction method under non-uniform illumination condition | |
| CN110687122A (en) | Method and system for detecting surface cracks of ceramic tile | |
| CN112184725B (en) | Method for extracting center of structured light bar of asphalt pavement image | |
| CN109815961B (en) | Pavement repairing type disease detection method based on local texture binary pattern | |
| CN105701490A (en) | Container number adaptive positioning method based on image entropy | |
| CN103914840B (en) | A kind of human body contour outline extraction method for non-simple background | |
| CN107610094B (en) | Three-dimensional crack detection method based on ellipsoid three-dimensional representation | |
| CN113763404B (en) | Foam image segmentation method based on optimization mark and edge constraint watershed algorithm | |
| CN108629776B (en) | Mineral rock granularity detecting system | |
| CN107886493A (en) | A kind of wire share split defect inspection method of transmission line of electricity | |
| KR102392252B1 (en) | Mushroom Object Growth Monitoring Method Using Image Processing Technology | |
| Hua et al. | Image segmentation algorithm based on improved visual attention model and region growing | |
| CN105787955A (en) | Sparse segmentation method and device of strip steel defect | |
| CN107239761A (en) | Fruit tree branch pulling effect evaluation method based on skeleton Corner Detection | |
| CN115330752A (en) | Pavement recessive stripping point detection method and system based on depth image recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |