CN112184725B - Method for extracting center of structured light bar of asphalt pavement image - Google Patents
Method for extracting center of structured light bar of asphalt pavement image Download PDFInfo
- Publication number
- CN112184725B CN112184725B CN202010992668.4A CN202010992668A CN112184725B CN 112184725 B CN112184725 B CN 112184725B CN 202010992668 A CN202010992668 A CN 202010992668A CN 112184725 B CN112184725 B CN 112184725B
- Authority
- CN
- China
- Prior art keywords
- light bar
- point
- center
- asphalt pavement
- pavement image
- 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.)
- Active
Links
- 239000010426 asphalt Substances 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000000605 extraction Methods 0.000 claims abstract description 19
- 238000001514 detection method Methods 0.000 claims abstract description 7
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 5
- 238000006243 chemical reaction Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 abstract description 2
- 239000000284 extract Substances 0.000 abstract description 2
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 210000004204 blood vessel Anatomy 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- 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
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- 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
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Geometry (AREA)
- Multimedia (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a method for extracting the center of a structured light bar of an asphalt pavement image, which comprises the following steps: dividing an image by using a light bar gray threshold, cutting a region of interest (ROI) of the image light bar, and carrying out a subsequent extraction algorithm in the ROI; adopting a non-deviation extraction method of the center of a Steger curve stripe to obtain a sub-pixel center point of the light bar; and (3) using a local outlier factor detection method LOF for the center point of the light bar extracted by the Steger, searching for an outlier by calculating the local density of each point, and removing to obtain an accurate center point of the light bar. The method for extracting the structured light bar center of the asphalt pavement image provided by the invention obviously improves the light bar center extraction effect, removes interference aiming at the characteristics of noise points, avoids unnecessary convolution calculation, has higher calculation efficiency, stronger universality and robustness, extracts the accurate light bar center so as to facilitate coordinate conversion of the center point, is applied to three-dimensional scanning and apparent reconstruction of the pavement surface, and improves the extraction precision and recognition efficiency of the apparent information of the asphalt pavement.
Description
Technical Field
The invention relates to a method for extracting a structured light bar center of an asphalt pavement image, belonging to the technical fields of road engineering and image processing.
Background
By the end of 2019, national highway mileage has reached 501.25 kilometers, with highway total mileage having reached 14.96 kilometers. As the most basic and focused foundation construction engineering with larger investment, the quality of the highway influences the economic development of various places and the daily living standard of people, so the detection and maintenance management of the highway increasingly shows importance and urgency. At present, with the development of industrial cameras, image processing and other technologies and the improvement of computing power, pavement three-dimensional scanning acquisition equipment is approaching to maturity, and the detection of the construction depth still adopts a completely manual mode such as a sand paving method, so that the workload is high and the efficiency is low. Aiming at a road surface structured light image development algorithm acquired by self-developed road surface structured light three-dimensional scanning equipment, a key technology of accurately and rapidly extracting three-dimensional scanning is performed on a light bar center, and a road surface apparent three-dimensional image can be accurately reconstructed after coordinate conversion is performed on a center point, so that road surface conditions can be timely, efficiently and accurately evaluated, and human resources are saved.
The current common method for extracting the center of the structured light bar mainly comprises the following steps:
(1) The traditional light bar center extraction method mainly comprises the following steps: extremum method, threshold method, gravity center method, curve fitting method, edge method, geometric center method, etc. these methods pass gray curve fitting in the fringe normal direction according to fringe gray distribution characteristics, the extracted center point is pixel level coordinate, the error is great, the precision is lower, are influenced by factors such as environment, equipment itself and the complexity of the measured object, these light stripe center extraction methods are difficult to satisfy comparatively general, real-time, accurate measurement requirement simultaneously.
(2) The central extraction method of the sub-pixel level is most typically a 'curve stripe center unbiased extraction method' proposed by Steger doctor, aiming at stripe (such as blood vessels, roads and the like) characteristics in medical images and satellite images, the sub-pixel center of gray distribution is obtained by expanding according to a Taylor polynomial, so that the method has stronger universality and robustness, and higher extraction precision, but because the method uses Gaussian kernel image convolution of a large template, the operation efficiency is lower, and noise is generated when the method is applied to images with more complex gray on asphalt pavement.
In the existing methods, the traditional method has higher requirements on road surface image quality, is greatly interfered by external environment, but asphalt pavement detection in actual engineering is not obtained under the same illumination condition (day/night), (sunny day/cloudy), the Steger method has obvious precision advantages, but the large-range Gaussian convolution calculation efficiency is low, noise is generated at the place with larger local gray level at the edge of a light bar, and the efficiency and accuracy of center extraction are affected.
Disclosure of Invention
In order to avoid the defects of low calculation efficiency, noise and the like of a Steger method, the invention improves a Steger stripe center extraction method by dividing light stripes and removing outliers, and extracts light stripe centers of asphalt pavement structure light images in batches so as to be better applied to pavement three-dimensional reconstruction.
The invention adopts the following technical scheme for solving the technical problems:
a method for extracting the center of a structured light bar of an asphalt pavement image comprises the following steps:
step 1, performing gray level binarization on an asphalt pavement image, dividing the light bar by threshold segmentation processing by utilizing the characteristic of the maximum integral gray level of the light bar in the asphalt pavement image, and expanding a certain distance outwards from the edge of the light bar to obtain an image light bar region of interest (ROI);
step 2, carrying out unbiased extraction of the center of the Steger curve stripe on the ROI to obtain a light bar sub-pixel center point;
and step 3, LOF outlier factor detection is carried out on the obtained sub-pixel center points of the light bar, outliers with smaller local density are removed, and the accurate center points of the light bar are obtained.
Further, the step 1 specifically includes the following steps:
step 1.1: after the grey level of the asphalt pavement image is changed, the grey level of the asphalt pavement image is changed into two values by using an im2bw function, and a grey level threshold I=0.98 is set according to the characteristic of large grey level of the light bar;
step 1.2: traversing the binarized asphalt pavement image, respectively searching points with pixel values of 1 from the upper, lower, left and right sides of the binarized asphalt pavement image, stopping after finding the first point, and recording row and column coordinates of the point, wherein row coordinates of the upper point and the lower point are the maximum value r of row coordinates of a light bar min Minimum value r max The coordinates of the left and right dot columns are the maximum value l of the coordinates of the light bar column min Minimum value l max ；
Step 1.3: calculating the ROI of the light bar, wherein the row coordinate range of the ROI is [ r ] min -a,r max +a]The column coordinate range is [ l ] min -a,l max +a]A is a preset value.
Further, the step 2 specifically includes the following steps:
step 2.1: convolving the ROI obtained in the step 1 with a two-dimensional Gaussian differential kernel to obtain a Hessian matrix H (x, y), and calculating a feature vector corresponding to the maximum feature value of the H (x, y), namely, a local normal direction of the stripe;
step 2.2: describing local gray distribution in the local normal direction of the stripe by adopting a second-order Taylor polynomial: p (x) =r+r' x+0.5r "x 2 Where r is a value of the second order Taylor polynomial developed at x=0, x is a pixel point in the direction of the local normal of the stripe, and a point p' (x) =0 is a sub-pixel center point of the stripe.
Further, the step 3 specifically includes the following steps:
step 3.1: calculating the local reachable density of the sub-pixel center point of the light bar obtained in the step 2Where point P is the kth point nearest to point O, N k (O) is the kth distance neighborhood of point O, d k (O, P) is the kth reachable distance from point P to point O, thereby calculating the local outlier factor +.>
Step 3.2: from local outlier factors LOF k And (O) the distribution, namely distinguishing abnormal points according to a set threshold value, and removing the abnormal points to obtain an accurate light bar center.
Compared with the prior art, the technical scheme provided by the invention has the following technical effects:
according to the characteristics, the invention aims at improving the defects of a deviation-free extraction method of the curve stripe center, effectively eliminating noise interference, extracting accurate light bar center information and improving the calculation efficiency;
the improved method for extracting the light bar center of the structural light bar is used for extracting the light bar center of the structural light image of the asphalt pavement, the algorithm can effectively eliminate the edge noise of the light bar, the operation efficiency is high, and the universality and the robustness are high.
Drawings
FIG. 1 is a road surface structured light image F acquired by a three-dimensional scanning device in an embodiment of the invention;
FIG. 2 is a region of interest ROI of a light bar segmented by a threshold in an embodiment of the present invention;
FIG. 3 is a graph showing the extraction of centerlines and determination of outliers by the Steger method in accordance with an embodiment of the present invention;
FIG. 4 is a bar center line after outlier removal in an embodiment of the present invention;
fig. 5 is a flow chart of the method of the present invention.
Detailed Description
The technical scheme of the invention is further described in detail below with reference to the accompanying drawings:
the present embodiment is based on the following assumption to achieve the extraction of the structured light stripe center of an asphalt pavement image:
1. the shooting range of the industrial camera is large, and the light bar area is only a small part;
2. the shooting edge of the camera generates distortion, and the brightness of the light bar is reduced;
3. the gray level of the structured light bar is the largest in the image;
4. the light bar is a narrow, continuous target object;
based on the above assumptions, the difference of the structured light bar from the rest of the image is exploited. As shown in fig. 5, the present embodiment divides an image by using a light bar gray threshold, cuts out a region of interest (ROI) of the image light bar, and performs a subsequent extraction algorithm in the ROI; adopting a non-deviation extraction method of the center of a Steger curve stripe to obtain a sub-pixel center point of the light bar; and (3) using a local outlier factor detection method (LOF) for the center point of the light bar extracted by the Steger, and obtaining an accurate center point of the light bar by calculating the local density of each point to find and remove the outlier. The embodiment does not depend on high-quality structured light images, and has good adaptability to complex pavement background and uneven illumination.
The embodiment specifically comprises the following steps:
step 1.1: after the image shown in fig. 1 is grayed, an im2bw function is used for binarizing the gray level image, the gray level threshold value is selected to be I=0.98 according to the characteristic of large gray level of the light bar, and only the light bar part is reserved in the binary image;
step 1.2: traversing the binarized asphalt pavement image, respectively searching points with pixel values of 1 from the upper, lower, left and right sides of the binarized asphalt pavement image, stopping after finding the first point, and recording row and column coordinates of the point, wherein row coordinates of the upper point and the lower point are the maximum value r of row coordinates of a light bar min Minimum value r max The coordinates of the left and right dot columns are the maximum value l of the coordinates of the light bar column min Minimum value l max ；
Step 1.3: calculating a light bar region of interest (ROI), r being in the range [ r ] min -10,r max +10]L ranges from [ l ] min -10,l max +10]Cutting out an ROI image from the original image as shown in figure 2;
step 2, performing a non-offset extraction method of the center of the curve stripe on the Region (ROI) where the light bar is positioned to obtain a sub-pixel center point of the light bar;
step 2.1: convolving the ROI image obtained in the step 1 with a two-dimensional Gaussian differential kernel to obtain a Hessian matrix H (x, y), and calculating a feature vector corresponding to the maximum feature value of the H (x, y), namely, a local normal direction of the stripe;
step 2.2: describing local gray distribution in the direction of the fringe normal by adopting a second-order Taylor polynomial, wherein p (x) =r+r 'x+0.5r' x 2 The point of p' (x) =0 is the center point of the sub-pixel of the light bar;
and step 3, detecting outlier factors of the obtained sub-pixel center points of the light bar, removing outlier points with smaller local density, and obtaining an accurate light bar center point.
Step 3.1: calculating the local reachable density of the sub-pixel center point of the light bar obtained in the step 2, selecting a k value of 5,thereby calculating the local outlier factor +.>
Step 3.2: from local outlier factors LOF k (O) distribution, selecting local outlier factor LOF k (O)>The 0.3 point is an outlier and is rejected, obtaining an accurate light bar center, as shown in fig. 3 and 4.
The foregoing is merely illustrative of the embodiments of the present invention, and the scope of the present invention is not limited thereto, and any person skilled in the art will appreciate that modifications and substitutions are within the scope of the present invention, and the scope of the present invention is defined by the appended claims.
Claims (3)
1. The method for extracting the center of the structured light bar of the asphalt pavement image is characterized by comprising the following steps of:
step 1, performing gray level binarization on an asphalt pavement image, dividing the light bar by threshold segmentation processing by utilizing the characteristic of the maximum integral gray level of the light bar in the asphalt pavement image, and expanding a certain distance outwards from the edge of the light bar to obtain an image light bar region of interest (ROI);
step 2, carrying out unbiased extraction of the center of the Steger curve stripe on the ROI to obtain a light bar sub-pixel center point;
step 3, LOF outlier factor detection is carried out on the obtained sub-pixel center points of the light bar, outliers with smaller local density are removed, and accurate center points of the light bar are obtained;
the step 3 specifically comprises the following steps:
step 3.1: calculating the local reachable density of the sub-pixel center point of the light bar obtained in the step 2Where point P is the kth point nearest to point O, N k (O) is the kth distance neighborhood of point O, d k (O, P) is the kth reachable distance from point P to point O, thereby calculating the local outlier factor +.>
Step 3.2: from local outlier factors LOF k And (O) the distribution, namely distinguishing abnormal points according to a set threshold value, and removing the abnormal points to obtain an accurate light bar center.
2. The method for extracting the center of a structured light bar of an asphalt pavement image according to claim 1, wherein the step 1 specifically comprises the steps of:
step 1.1: after the grey level of the asphalt pavement image is changed, the grey level of the asphalt pavement image is changed into two values by using an im2bw function, and a grey level threshold I=0.98 is set according to the characteristic of large grey level of the light bar;
step 1.2: traversing the binarized asphalt pavement image, respectively searching points with pixel values of 1 from the upper, lower, left and right sides of the binarized asphalt pavement image, stopping after finding the first point, and recording row and column coordinates of the point, wherein row coordinates of the upper point and the lower point are the maximum value r of row coordinates of a light bar min Minimum value r max The coordinates of the left and right dot columns are the maximum value l of the coordinates of the light bar column min Minimum value l max ；
Step 1.3: calculating the ROI of the light bar, wherein the row coordinate range of the ROI is [ r ] min -a,r max +a]The column coordinate range is [ l ] min -a,l max +a]A is a preset value.
3. The method for extracting the center of a structured light bar of an asphalt pavement image according to claim 1, wherein the step 2 specifically comprises the steps of:
step 2.1: convolving the ROI obtained in the step 1 with a two-dimensional Gaussian differential kernel to obtain a Hessian matrix H (x, y), and calculating a feature vector corresponding to the maximum feature value of the H (x, y), namely, a local normal direction of the stripe;
step 2.2: describing local gray distribution in the local normal direction of the stripe by adopting a second-order Taylor polynomial: p (x) =r+r' x+0.5r "x 2 Where r is a value of the second order Taylor polynomial developed at x=0, x is a pixel point in the direction of the local normal of the stripe, and a point p' (x) =0 is a sub-pixel center point of the stripe.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010992668.4A CN112184725B (en) | 2020-09-21 | 2020-09-21 | Method for extracting center of structured light bar of asphalt pavement image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010992668.4A CN112184725B (en) | 2020-09-21 | 2020-09-21 | Method for extracting center of structured light bar of asphalt pavement image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112184725A CN112184725A (en) | 2021-01-05 |
| CN112184725B true CN112184725B (en) | 2024-02-02 |
Family
ID=73957116
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010992668.4A Active CN112184725B (en) | 2020-09-21 | 2020-09-21 | Method for extracting center of structured light bar of asphalt pavement image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112184725B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113129357B (en) * | 2021-05-10 | 2022-09-30 | 合肥工业大学 | Method for extracting light strip center in three-dimensional scanning measurement under complex background |
| CN115423807B (en) * | 2022-11-04 | 2023-03-24 | 山东益民服饰有限公司 | Cloth defect detection method based on outlier detection |
| CN116433707B (en) * | 2023-06-14 | 2023-08-11 | 武汉工程大学 | Accurate extraction method and system for optical center sub-pixels of line structure under complex background |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101178812A (en) * | 2007-12-10 | 2008-05-14 | 北京航空航天大学 | Mixed image processing process of structure light striation central line extraction |
| WO2012100522A1 (en) * | 2011-01-26 | 2012-08-02 | 南京大学 | Ptz video visibility detection method based on luminance characteristic |
-
2020
- 2020-09-21 CN CN202010992668.4A patent/CN112184725B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101178812A (en) * | 2007-12-10 | 2008-05-14 | 北京航空航天大学 | Mixed image processing process of structure light striation central line extraction |
| WO2012100522A1 (en) * | 2011-01-26 | 2012-08-02 | 南京大学 | Ptz video visibility detection method based on luminance characteristic |
Non-Patent Citations (1)
| Title |
|---|
| 结构光光条提取的混合图像处理方法;周富强;陈强;张广军;;光电子.激光(第11期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112184725A (en) | 2021-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112184725B (en) | Method for extracting center of structured light bar of asphalt pavement image | |
| Cheng et al. | 3D building model reconstruction from multi-view aerial imagery and lidar data | |
| CN112991347B (en) | Three-dimensional-based train bolt looseness detection method | |
| CN106067003B (en) | Automatic extraction method for road vector identification line in vehicle-mounted laser scanning point cloud | |
| CN106530281B (en) | Unmanned plane image fuzzy Judgment method and system based on edge feature | |
| CN110675392B (en) | Rut fine three-dimensional feature extraction method based on pavement continuous laser point cloud | |
| CN112131963B (en) | Road identification line extraction method based on driving direction structural feature constraint | |
| CN110334692B (en) | Blind road identification method based on image processing | |
| CN103793708A (en) | Multi-scale license plate precise locating method based on affine correction | |
| CN107895151A (en) | Method for detecting lane lines based on machine vision under a kind of high light conditions | |
| CN108921076B (en) | Pavement crack disease self-adaptive constant false alarm detection method based on image | |
| CN112633092A (en) | Road information extraction method based on vehicle-mounted laser scanning point cloud | |
| CN111968079B (en) | Three-dimensional pavement crack extraction method based on local extremum of section and segmentation sparsity | |
| CN112669333A (en) | Single tree information extraction method | |
| CN114820474A (en) | Train wheel defect detection method based on three-dimensional information | |
| CN113129323A (en) | Remote sensing ridge boundary detection method and system based on artificial intelligence, computer equipment and storage medium | |
| CN110211128B (en) | Loess plateau terrace extraction method based on remote sensing image and DEM | |
| CN112330667B (en) | Morphology-based laser stripe center line extraction method | |
| CN106324708B (en) | Digitizing solution, the device of rainfall record drawing | |
| CN107578414B (en) | Method for processing pavement crack image | |
| CN103065296B (en) | High-resolution remote sensing image residential area extraction method based on edge feature | |
| CN112561808A (en) | Road boundary line restoration method based on vehicle-mounted laser point cloud and satellite image | |
| CN113066041A (en) | Pavement crack detection method based on stack sparse self-coding deep learning | |
| CN116416589A (en) | Rail point cloud extraction method based on rail data and geometric properties | |
| Bao | Road distress analysis using 2D and 3D information |
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 |