CN110322441A - The visible detection method of the apparent slight crack in road surface - Google Patents
The visible detection method of the apparent slight crack in road surface Download PDFInfo
- Publication number
- CN110322441A CN110322441A CN201910612772.3A CN201910612772A CN110322441A CN 110322441 A CN110322441 A CN 110322441A CN 201910612772 A CN201910612772 A CN 201910612772A CN 110322441 A CN110322441 A CN 110322441A
- Authority
- CN
- China
- Prior art keywords
- image
- slight crack
- road surface
- block
- calculated
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 37
- 230000002708 enhancing effect Effects 0.000 claims abstract description 17
- 238000001914 filtration Methods 0.000 claims abstract description 13
- 230000011218 segmentation Effects 0.000 claims abstract description 8
- 230000000877 morphologic effect Effects 0.000 claims abstract description 6
- 238000012360 testing method Methods 0.000 claims abstract description 3
- 238000009792 diffusion process Methods 0.000 claims description 12
- 238000000034 method Methods 0.000 claims description 10
- 238000010606 normalization Methods 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 9
- 239000007787 solid Substances 0.000 claims description 5
- MTEOMEWVDVPTNN-UHFFFAOYSA-E almagate Chemical compound O.O.[OH-].[OH-].[OH-].[OH-].[OH-].[OH-].[OH-].[Mg+2].[Mg+2].[Mg+2].[Al+3].[O-]C([O-])=O MTEOMEWVDVPTNN-UHFFFAOYSA-E 0.000 claims description 4
- 238000013100 final test Methods 0.000 claims description 3
- 230000002452 interceptive effect Effects 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000003796 beauty Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000003014 reinforcing effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30132—Masonry; Concrete
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to the visible detection methods of the apparent slight crack in road surface, comprising: A. input is to be detected to contain crackled pavement image I (x, y)；B. 4 directions in road pavement image I (x, y) around each pixel are filtered by diffusing filter algorithm, obtain smoothed image；C. by Kuwahara filtering algorithm, the smoothed image is filtered, low noise image is obtained；D. the slight crack region in low noise image is enhanced by Jerman filtering algorithm, obtains enhancing image；E. the bianry image to enhancing image by Double Thresholding Segmentation algorithm, after being divided；F. the image block in bianry image is screened according to predetermined fixed threshold using morphological properties, the testing result after being screened.The present invention can significantly improve the robustness of road surface crack detection method and the accuracy rate of detection, considerably reduce false detection rate.
Description
Technical field
The present invention relates to the image processing methods of machine vision applications, are especially the vision-based detection side of the apparent slight crack in road surface
Method.
Background technique
Since highway will receive various damages, vehicle and pedestrian travelled on highway is caused more or less to exist
Certain security risk, while damaged highway also very influences beauty, this is but also highway slight crack detection technique becomes highway
Very important project among maintenance.Since the detection system of view-based access control model has the advantages such as precision is high, signal-to-noise ratio is good,
Domestic and international many researchers are just energetically applying it among the detection of highway slight crack.But slight crack signal is presented pair in the picture
Feature lower than degree and tiny, directly detection are very difficult, therefore how slight crack letter are accurately detected from road surface image
Number become the hot spot of Recent study.
Image enhancement technique is the key that small-signal in detection image, image enhancement be intended to enhance needed in image it is weak
Small signal, and inhibit unwanted interference.So the effect quality of image enhancement directly affects final detection accuracy and effect
Fruit.
In the road surface slight crack detection technique of current view-based access control model, has a large amount of research achievement.Not excessive portion
The research divided is all the slight crack region detected in road surface image using the method for characteristic point combination machine learning, these sides
Method needs the data for being largely used to train classifier, and the case where practical highway slight crack is changeable and many kinds of, and data is caused to be received
Collection is not easy and algorithm robustness is not high, and Detection accuracy is lower, and erroneous detection is higher.
Summary of the invention
The present invention provides a kind of visible detection methods of the apparent slight crack in road surface, improve road surface crack detection method
The accuracy rate of robustness and detection reduces false detection rate.
The visible detection method of the apparent slight crack in road surface of the present invention, comprising:
A. it inputs to be detected containing crackled pavement image I (x, y)；
B. 4 directions in road pavement image I (x, y) around each pixel are filtered by diffusing filter algorithm, are obtained
To smoothed image；
C. by Kuwahara filtering algorithm, the smoothed image is filtered, low noise image is obtained；
D. the slight crack region in low noise image is enhanced by Jerman filtering algorithm, obtains enhancing image；
E. the bianry image to enhancing image by Double Thresholding Segmentation algorithm, after being divided；
F. the image block in bianry image is screened according to predetermined fixed threshold using morphological properties (Gu
Determining threshold value can choose based on experience value), the testing result after being screened.
Further, in step B, the gradient in 4 directions around each pixel in pavement image I (x, y) is first calculated, then
The diffusion coefficient that 4 directions are calculated according to gradient, finally calculates filter result according to gradient and diffusion coefficient, to obtain
Smoothed image.
Specifically, step B includes:
B1. the gradient in 4 directions around each pixel in pavement image I (x, y) is first calculated:
Wherein, x and y respectively indicates the abscissa and ordinate of current pixel,Respectively indicate the gradient in 4 directions around current pixel；
B2. the diffusion coefficient in 4 directions of pixel is calculated:
Wherein, K is the coefficient of heat conduction on road surface, cNx,y,cSx,y,cEx,y,cWx,yIt respectively indicates around pixel on 4 directions
Diffusion coefficient；
B3. filter result is calculated according to gradient and diffusion coefficient:
Wherein, t is the number of iterations, and it is weight coefficient that rule of thumb general value, which is 5, λ, is typically set to 0.25；
B4. step B1 to B3 is repeated, until completing the number of iterations t.
Further, smoothed image is filtered described in step C, is first taken out in pavement image I (x, y) with every
Point, the image block A that size is z × z, are then four subimage blocks by block image A points, finally calculate separately centered on a pixel
The variance of this four subimage blocks takes result of the mean value in the wherein the smallest region of variance as the central point pixel filter.
Specifically, wherein described divide block image A for four subimage blocks, comprising:
The size of image block A is z × z, and the 1st of image block A is taken to arrange to theColumn, the 1st row to theCapable data
For the first subimage block a1；
Take the of image block AIt arranges to z column, the 1st row to theCapable data are the second subimage block a2；
The 1st of image block A is taken to arrange to theColumn, theThe data of row to z row are third subimage block a3；
Take the of image block AIt arranges to z column, theThe data of row to z row are the 4th subimage block a4；
If calculated result is decimal, a position is rounded up to the fractional part of the calculated result.
Further, the slight crack region in low noise image, packet are enhanced by Jerman filtering algorithm described in step D
Include: calculate pavement image I (x, y) Hessian matrix, then calculate Hessian matrix characteristic value, to the characteristic value into
Row normalized obtains normalization characteristic value, and the enhancing figure is finally calculated according to characteristic value and normalization characteristic value
Picture.
Specifically, step D includes:
D1. the Hessian matrix of pavement image I (x, y) is calculated；
D2. the eigenvalue λ of Hessian matrix is calculated2；
D3. to eigenvalue λ2It is normalized:
Wherein, λρFor normalization characteristic value, s is scale factor, s can between 0.5~2 value, rule of thumb generally
Value be 1, τ be 0~1 between interceptive value, generally according to experience value be 1, λ2(x, s) is indicated at x and having a size of s's
Characteristic value, wherein x=[x1,x2] indicates coordinate vector, x1,x2Respectively abscissa and ordinate；
D4. according to normalization characteristic value λρAnd eigenvalue λ2, calculate enhancing image VJ:
Specifically, step E includes:
E1. two fixed threshold value t are set1And t2, it is generally set to 0.5 and 0.9 and superposition image prime number p, is generally set to
4；
E2. pass through two fixed threshold value t1And t2Threshold process is carried out to enhancing image respectively, respectively obtains corresponding two
It is worth image BW1And BW2；
E3. according to bianry image BW1And BW2, calculate the region superposition image prime number in each regionIt is heavy when meeting region
Fold-over prime numberWithout label when being marked, and be unsatisfactory for the region when > superposition image prime number p；
E4. it is exported all containing markd region as segmentation result.
Specifically, step F includes:
F1. the non-zero area attribute of image block in the bianry image of input, the non-zero region are calculated using morphological operation
Attribute includes: long axis length ALmax, minor axis length ALmin, central point and density solid, wherein the long axis length ALmax
With minor axis length ALminCalculation may refer to: Haralick R M, Shapiro L G.Computer and robot
Vision [M] .Reading:Addison-wesley, 1992, Appendix A；
The calculation of central point are as follows: cross, the ordinate in region non-zero in each image block are calculated separately into average value,
The as center position in the non-zero region of the image block；
The calculation of density solid are as follows: the sum of all pixels in the non-zero region of each image block is minimum convex more divided by it
The non-zero sum of all pixels of side shape, wherein the calculation of minimal convex polygon may refer to: Gonzalez R C, Woods R
E.Digital image processing[M][J].Publishing house of electronics industry,
2002,141(7):416；
F2. judge whether described image block meets the attribute being calculated and want using the predetermined fixed threshold
It asks, retains the image block if meeting, otherwise remove the image block, obtain final testing result.
The beneficial effect of the visible detection method of the apparent slight crack in road surface of the present invention includes:
1, the tiny slight crack signal in pavement image is enhanced by Jerman filtering algorithm, hence it is evident that improve slight crack signal
Contrast, while by the way that by a large amount of experiment discovery, this reinforcing effect has good robustness and fits to scene
Ying Xing greatly improves the robustness and detection accuracy of road surface slight crack detection algorithm.
2, it in conjunction with the characteristics of anisotropic filter and Kuwahara filter, has effective filtered out in original pavement image
Random noise, while also remaining faint slight crack signal as far as possible, effectively increase slight crack detection algorithm robustness and
Anti-interference ability.
3, method calculating process of the invention is easy, to the adaptable of various backgrounds.
Specific embodiment with reference to embodiments is described in further detail above content of the invention again.
But the range that this should not be interpreted as to the above-mentioned theme of the present invention is only limitted to example below.Think not departing from the above-mentioned technology of the present invention
In the case of thinking, the various replacements or change made according to ordinary skill knowledge and customary means should all be included in this hair
In bright range.
Detailed description of the invention
Fig. 1 is the flow chart of the visible detection method of the apparent slight crack in road surface of the present invention.
Fig. 2 is the pavement image inputted in embodiment.
Fig. 3 is the low noise image that Fig. 2 is obtained by Kuwahara filtering algorithm.
Fig. 4 is that Fig. 3 obtains enhancing image by Jerman filtering algorithm.
Fig. 5 is bianry image of the Fig. 4 after Double Thresholding Segmentation.
Fig. 6 is finally obtained slight crack detection result image.
Specific embodiment
The visible detection method of the apparent slight crack in road surface of the present invention as shown in Figure 1, comprising:
A. as shown in Fig. 2, input is to be detected to contain crackled pavement image I (x, y)；
B. 4 directions in road pavement image I (x, y) around each pixel are filtered by diffusing filter algorithm, are obtained
To smoothed image, comprising:
B1. the gradient in 4 directions around each pixel in pavement image I (x, y) is first calculated:
Wherein, x and y respectively indicates the abscissa and ordinate of current pixel,Respectively indicate the gradient in 4 directions around current pixel；
B2. the diffusion coefficient in 4 directions of pixel is calculated:
Wherein, K is the coefficient of heat conduction on road surface, cNx,y,cSx,y,cEx,y,cWx,yIt respectively indicates around pixel on 4 directions
Diffusion coefficient；
B3. filter result is calculated according to gradient and diffusion coefficient:
Wherein, t is the number of iterations, and it is weight coefficient that value, which is 5, λ, in the present embodiment, and value is 0.25 in the present embodiment；
B4. step B1 to B3 is repeated, until completing the number of iterations t.
C. by Kuwahara filtering algorithm, the smoothed image is filtered, low noise image is obtained, wrapped
It includes:
C1. first take out pavement image I (x, y) in centered on each pixel point, size for z × z image block A.
C2. it is four subimage blocks by block image A points:
The 1st of image block A is taken to arrange to theColumn, the 1st row to theCapable data are the first subimage block a1；
Take the of image block AIt arranges to z column, the 1st row to theCapable data are the second subimage block a2；
The 1st of image block A is taken to arrange to theColumn, theThe data of row to z row are third subimage block a3；
Take the of image block AIt arranges to z column, theThe data of row to z row are the 4th subimage block a4；
If calculated result is decimal, a position is rounded up to the fractional part of the calculated result.
C3. the variance for calculating separately this four subimage blocks takes the mean value in the wherein the smallest region of variance as the center
The result of point pixel filter.All filtered pixels constitute low noise image as shown in Figure 3.
D. the slight crack region in low noise image is enhanced by Jerman filtering algorithm, obtains enhancing image, specifically:
D1. the Hessian matrix of pavement image I (x, y) is calculated.
D2. the eigenvalue λ of Hessian matrix is calculated2；
D3. to eigenvalue λ2It is normalized:
Wherein, λρFor normalization characteristic value, s is scale factor, and s=1 in the present embodiment, τ are the truncation threshold between 0~1
It is worth, τ=1, λ in the present embodiment2(x, s) is indicated at x and the characteristic value having a size of s, wherein x=[x1,x2] indicates coordinate to
Amount, x1,x2Respectively abscissa and ordinate；
D4. according to normalization characteristic value λρAnd eigenvalue λ2, calculate enhancing image VJ, as shown in Figure 4:
E. Double Thresholding Segmentation algorithm is passed through to enhancing image, the bianry image after being divided, specifically:
E1. two fixed threshold value t are set1And t2, respectively 0.5 and 0.9 and superposition image prime number p=4；
E2. pass through two fixed threshold value t1And t2Threshold process is carried out to enhancing image respectively, respectively obtains corresponding two
It is worth image BW1And BW2；
E3. to the bianry image BW1And BW2Progress and operation, then seek pixel value all in result images
With calculate the region superposition image prime number in each regionWhen meeting region superposition image prime numberWhen > superposition image prime number p pair
The region is marked, and without label when be unsatisfactory for；
E4. it is exported all containing markd region as segmentation result, as shown in Figure 5.
F. the non-zero area attribute of image block in the bianry image of input is calculated using morphological operation, the non-zero region belongs to
Property includes: long axis length ALmax, minor axis length ALmin, central point and density solid.Wherein, the long axis length ALmaxWith
Minor axis length ALminCalculation may refer to: Haralick R M, Shapiro L G.Computer and robot
Vision [M] .Reading:Addison-wesley, 1992, Appendix A；
The calculation of central point are as follows: cross, the ordinate in region non-zero in each image block are calculated separately into average value,
The as center position in the non-zero region of the image block；
The calculation of density solid are as follows: the sum of all pixels in the non-zero region of each image block is minimum convex more divided by it
The non-zero sum of all pixels of side shape, wherein the calculation of minimal convex polygon may refer to: Gonzalez R C, Woods R
E.Digital image processing[M][J].Publishing house of electronics industry,
2002,141(7):416。
Then fixed threshold is determined by empirical value, judges whether described image block meets meter using determining fixed threshold
Obtained attribute specification retains the image block if meeting, otherwise removes the image block, obtains final testing result,
As shown in fig. 6, wherein slight crack region is marked in Fig. 6.
Claims (9)
1. the visible detection method of the apparent slight crack in road surface, feature include:
A. it inputs to be detected containing crackled pavement image I (x, y)；
B. 4 directions in road pavement image I (x, y) around each pixel are filtered by diffusing filter algorithm, are put down
Sliding image；
C. by Kuwahara filtering algorithm, the smoothed image is filtered, low noise image is obtained；
D. the slight crack region in low noise image is enhanced by Jerman filtering algorithm, obtains enhancing image；
E. the bianry image to enhancing image by Double Thresholding Segmentation algorithm, after being divided；
F. the image block in bianry image is screened using morphological properties according to predetermined fixed threshold, is sieved
Testing result after choosing.
2. the visible detection method of the apparent slight crack in road surface as described in claim 1, it is characterized in that: in step B, first calculate outlet
In face image I (x, y) around each pixel 4 directions gradient, the diffusion coefficient in 4 directions is calculated further according to gradient, most
Filter result is calculated according to gradient and diffusion coefficient afterwards, to obtain smoothed image.
3. the visible detection method of the apparent slight crack in road surface as claimed in claim 2, it is characterized in that: step B includes:
B1. the gradient in 4 directions around each pixel in pavement image I (x, y) is first calculated:
Wherein, x and y respectively indicates the abscissa and ordinate of current pixel,
Respectively indicate the gradient in 4 directions around current pixel；
B2. the diffusion coefficient in 4 directions of pixel is calculated:
Wherein, K is the coefficient of heat conduction on road surface, cNx,y,cSx,y,cEx,y,cWx,yRespectively indicate the expansion around pixel on 4 directions
Dissipate coefficient；
B3. filter result is calculated according to gradient and diffusion coefficient:
Wherein, t is the number of iterations, and λ is weight coefficient；
B4. step B1 to B3 is repeated, until completing the number of iterations t.
4. the visible detection method of the apparent slight crack in road surface as described in claim 1, it is characterized in that: to smooth described in step C
Image is filtered, first take out pavement image I (x, y) in centered on each pixel point, size for z × z image block
A, wherein z is the odd number greater than 1 less than 9, is then four subimage blocks by block image A points, finally calculates separately this four sons
The variance of image block takes result of the mean value in the wherein the smallest region of variance as the central point pixel filter.
5. the visible detection method of the apparent slight crack in road surface as claimed in claim 4, it is characterized in that: it is wherein described by block image A
It is divided into four subimage blocks, comprising:
The size of image block A is z × z, and the 1st of image block A is taken to arrange to theColumn, the 1st row to theCapable data are the
One subimage block a1；
Take the of image block AIt arranges to z column, the 1st row to theCapable data are the second subimage block a2；
The 1st of image block A is taken to arrange to theColumn, theThe data of row to z row are third subimage block a3；
Take the of image block AIt arranges to z column, theThe data of row to z row are the 4th subimage block a4；
If calculated result is decimal, a position is rounded up to the fractional part of the calculated result.
6. the visible detection method of the apparent slight crack in road surface as described in claim 1, it is characterized in that: pass through described in step D
Jerman filtering algorithm enhances the slight crack region in low noise image, comprising: calculates the Hessian square of pavement image I (x, y)
Then battle array calculates the characteristic value of Hessian matrix, is normalized to the characteristic value, obtains normalization characteristic value, most
The enhancing image is calculated according to characteristic value and normalization characteristic value afterwards.
7. the visible detection method of the apparent slight crack in road surface as claimed in claim 6, it is characterized in that: step D includes:
D1. the Hessian matrix of pavement image I (x, y) is calculated；
D2. the eigenvalue λ of Hessian matrix is calculated2；
D3. to eigenvalue λ2It is normalized:
Wherein, λρFor normalization characteristic value, s is scale factor, and τ is the interceptive value between 0~1, λ2(x, s) indicate at x and
Characteristic value having a size of s, wherein x=[x1,x2] indicates coordinate vector, x1,x2Respectively abscissa and ordinate；
D4. according to normalization characteristic value λρAnd eigenvalue λ2, calculate enhancing image VJ:
8. the visible detection method of the apparent slight crack in road surface as described in claim 1, it is characterized in that: step E includes:
E1. two fixed threshold value t are set1And t2And superposition image prime number p；
E2. pass through two fixed threshold value t1And t2Threshold process is carried out to enhancing image respectively, respectively obtains corresponding binary map
As BW1And BW2；
E3. according to bianry image BW1And BW2, calculate the region superposition image prime number in each regionWhen meeting region superposition image
Prime numberWithout label when being marked, and be unsatisfactory for the region when > superposition image prime number p；
E4. it is exported all containing markd region as segmentation result.
9. the visible detection method of the apparent slight crack in road surface as described in claim 1, it is characterized in that: step F includes:
F1. the non-zero area attribute of image block in the bianry image of input, the non-zero area attribute are calculated using morphological operation
It include: long axis length ALmax, minor axis length ALmin, central point and density solid；
F2. judge whether described image block meets the attribute specification being calculated using the predetermined fixed threshold, if
Satisfaction then retains the image block, on the contrary then remove the image block, obtains final testing result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910612772.3A CN110322441A (en) | 2019-07-09 | 2019-07-09 | The visible detection method of the apparent slight crack in road surface |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910612772.3A CN110322441A (en) | 2019-07-09 | 2019-07-09 | The visible detection method of the apparent slight crack in road surface |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110322441A true CN110322441A (en) | 2019-10-11 |
Family
ID=68123106
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910612772.3A Pending CN110322441A (en) | 2019-07-09 | 2019-07-09 | The visible detection method of the apparent slight crack in road surface |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110322441A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112465817A (en) * | 2020-12-17 | 2021-03-09 | 大连海事大学 | Pavement crack detection method based on directional filter |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106802171A (en) * | 2017-03-22 | 2017-06-06 | 淄博贝林电子有限公司 | Shooting direct-reading water meter and automatic pollution removing method with automatic pollution removing function |
| CN108460744A (en) * | 2018-03-21 | 2018-08-28 | 武汉理工大学 | A kind of enhancing of cement cutting pavement image noise reduction and FRACTURE CHARACTERISTICS extracting method |
| CN109407149A (en) * | 2018-10-10 | 2019-03-01 | 电子科技大学 | Seismic coherence data crack detection method based on Hessian matrix |
-
2019
- 2019-07-09 CN CN201910612772.3A patent/CN110322441A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106802171A (en) * | 2017-03-22 | 2017-06-06 | 淄博贝林电子有限公司 | Shooting direct-reading water meter and automatic pollution removing method with automatic pollution removing function |
| CN108460744A (en) * | 2018-03-21 | 2018-08-28 | 武汉理工大学 | A kind of enhancing of cement cutting pavement image noise reduction and FRACTURE CHARACTERISTICS extracting method |
| CN109407149A (en) * | 2018-10-10 | 2019-03-01 | 电子科技大学 | Seismic coherence data crack detection method based on Hessian matrix |
Non-Patent Citations (4)
| Title |
|---|
| PIETRO PERONA ET AL.: "Scale-Space and Edge Detection Using Anisotropic Diffusion", 《IEEE》 * |
| TIM JERMAN ET AL.: "Enhancement of Vascular Structures in 3D and 2D Angiographic Images", 《IEEE》 * |
| YI LUO ET AL.: "Edge-preserving smoothing and applications", 《THE LEADING EDGE》 * |
| 刘勇: "镜面目标表观裂纹视觉测量技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112465817A (en) * | 2020-12-17 | 2021-03-09 | 大连海事大学 | Pavement crack detection method based on directional filter |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107330376B (en) | Lane line identification method and system | |
| KR100780957B1 (en) | Apparatus and method for selecting image | |
| CN103632159B (en) | Method and system for training classifier and detecting text area in image | |
| Wang et al. | Multifocus image fusion using convolutional neural networks in the discrete wavelet transform domain | |
| Fan et al. | Real-time stereo vision-based lane detection system | |
| WO2007055359A1 (en) | Clustering system and image processing system having same | |
| CN108171695A (en) | A kind of express highway pavement detection method based on image procossing | |
| CN110009653A (en) | Increase limb recognition point sharp picture based on gray level threshold segmentation method and knows method for distinguishing | |
| CN110491132A (en) | Vehicle based on video frame picture analyzing, which is disobeyed, stops detection method and device | |
| CN112149458A (en) | Obstacle detection method, intelligent driving control method, device, medium, and apparatus | |
| WO2021003936A1 (en) | Image segmentation method, electronic device, and computer-readable storage medium | |
| Sahoo et al. | Design and simulation of various edge detection techniques using Matlab Simulink | |
| CN111696080A (en) | Face fraud detection method, system and storage medium based on static texture | |
| CN109190639A (en) | A kind of vehicle color identification method, apparatus and system | |
| Krishnan et al. | An efficient texture classification algorithm using integrated Discrete Wavelet Transform and local binary pattern features | |
| CN112288758B (en) | Infrared and visible light image registration method for power equipment | |
| US8395824B2 (en) | Method for determining ground line | |
| CN101630365B (en) | Method for extracting and describing DAISY-based feature with mirror face turning invariance | |
| CN114708566A (en) | Improved YOLOv 4-based automatic driving target detection method | |
| CN112488958A (en) | Image contrast enhancement method based on scale space | |
| CN110322441A (en) | The visible detection method of the apparent slight crack in road surface | |
| Dai et al. | Recognition of tunnel lining cracks based on digital image processing | |
| CN117392375A (en) | Target detection algorithm for tiny objects | |
| CN107832732A (en) | Method for detecting lane lines based on ternary tree traversal | |
| CN105913405B (en) | The processing method of image detail, device and vehicle for rendering |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191011 |
|
| RJ01 | Rejection of invention patent application after publication |