CN107463939B - Image key straight line detection method - Google Patents
Image key straight line detection method Download PDFInfo
- Publication number
- CN107463939B CN107463939B CN201710503951.4A CN201710503951A CN107463939B CN 107463939 B CN107463939 B CN 107463939B CN 201710503951 A CN201710503951 A CN 201710503951A CN 107463939 B CN107463939 B CN 107463939B
- Authority
- CN
- China
- Prior art keywords
- straight line
- curve
- point
- image
- straight
- 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.)
- Expired - Fee Related
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 40
- 238000000034 method Methods 0.000 claims abstract description 42
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 12
- 238000003708 edge detection Methods 0.000 claims description 6
- 238000001914 filtration Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 abstract description 4
- 238000004458 analytical method Methods 0.000 abstract description 3
- 230000009466 transformation Effects 0.000 description 7
- 238000012216 screening Methods 0.000 description 5
- 238000012937 correction Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 238000004883 computer application Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000005452 bending Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an image key straight line detection method which comprises three parts of image preprocessing, straight line detection and key straight line judgment. The method utilizes the splitting method and the least square method to carry out the line detection, the algorithm efficiency and the accuracy are high, meanwhile, the K-means algorithm is utilized to carry out the line clustering firstly, and then the key line of the image is found according to the total length of the category line and the length of the line, so that the accuracy of the line detection is effectively improved, and the method provides guarantee for the processes of further target positioning, analysis, identification and the like.
Description
Technical Field
The invention relates to a method for detecting key straight lines of an image, in particular to a method for detecting key straight lines of an image based on straight line detection and clustering.
Background
The image key straight line detection method is used for detecting the straight line with the largest proportion in the image, comprises three parts of image preprocessing, straight line detection and key straight line judgment, and provides conditions for the processes of further target positioning, analysis, identification and the like. At present, the algorithm of line detection is mature, and the method can be mainly divided into two types of line detection methods based on Hough transformation and line detection methods based on chain codes.
The line detection method based on Hough change is a global method, has the advantages of strong anti-interference capability and large calculated amount, and has poor line detection effect on the line with a certain bending phenomenon, and is easy to generate false detection and missing detection. Aiming at the defects of a line detection method based on Hough change, many scholars propose an improved Hough line detection method, such as Yunyan swallow (research [ J ] based on an RHT-LSM line detection method, 2007,34(01):55-58) which combines random Hough transformation and least square method to carry out line detection, the performance of the traditional Hough change method is improved by using random Hough transformation, and the capacity of detecting a micro-bent line by Hough transformation is improved by using the least square method, but the method has the premise that the Hough transformation can firstly locate the approximate position of a target line and then carry out correction by using the least square method, so the problem of missing detection of the Hough method is not solved; qiaoyin (random Hough transformation straight line detection [ J ] based on least square correction, computer application 2015,35(11): 3312) 3315) provides a random Hough transformation straight line detection method based on rho-theta domain minimum two-fit correction, which improves the conditions of false detection and missed detection of Hough straight line detection on a standard straight line, but cannot ensure the detection effect on a slightly bent straight line.
The method has the advantages that the straight line detection accuracy is high, but the method is easy to fall into local optimization, so that the found straight line is too short, and meanwhile, the algorithm complexity is high. For example, Luguangquan (a straight line segment detection method [ J ] based on chain code detection, computer application 2006,32(14):1-3) and Cuzhen (an edge tracking algorithm based on Freeman chain code and a straight line segment detection [ J ] microcomputer application 2008,24(1):17-20) find angular points meeting the segmentation requirements through iteration, and then perform curve segmentation, the angular point judgment methods of the two methods are different, and meanwhile, the problem that the chain code method is easy to fall into local optimization is also improved differently, but the algorithm complexity problem still exists.
The invention adopts a split-based line detection method, which is similar to a chain code-based line detection method and needs to find angular points for curve segmentation, and the difference is that the method does not need to judge the angular points of all points, so that the algorithm efficiency is better, and the problem of local optimization does not exist. On the basis, the invention clusters the detected straight lines according to the angles by an improved K-means clustering method and determines the key straight lines in the images according to a certain screening strategy.
Disclosure of Invention
In order to solve the technical problems in the prior art, the invention provides an image key line detection method. In the image preprocessing process, an edge image of a window corner image is obtained through filtering and a Canny edge detection method; in the process of line detection, firstly, a curve set in an image is obtained, then all curves are split through a splitting method to obtain a curve set meeting line constraint, and finally a line set is obtained through least square line fitting; in the key straight line screening process, firstly, all straight lines are clustered according to angles by using an improved K-means method, then, a straight line set with the longest total length of the straight lines is screened out to be a set in which the key straight lines are located, and a line segment with the longest length in the set is selected to be a final key straight line.
The invention specifically comprises the following steps:
step 1: zooming the original image into an image D with width and height respectively;
step 2: gaussian filtering and Canny edge detection are carried out on the image DOperation is carried out to obtain an edge map D1An image D1Is B ═ Bi|i＝0,1,…,NB-1}，biRepresenting an image D1At an edge point of, and point biThe abscissa and ordinate of (A) are denoted by bi.x,bi.y，NBRepresenting the number of elements in set B;
and step 3: obtaining an image D1The curve set C comprises the following specific steps:
step 3.1: if there is a sufficient PN (B) in set Bk)∈{20,21,…,27,20+21,21+22,…,27+20Point b ofkThen b iskIs determined as a new curve CzStarting point of (1), initializing curve Cz＝{bkWill point bkRemoving from the set B, and turning to the step 3.2; otherwise, turning to the step 4; wherein PN (b)k) Represents point bkThe eight neighborhood values are obtained by calculation according to a formula (1);
wherein, E (x, y) represents whether a point with coordinates (x, y) exists in the set B, if so, E (x, y) is 1, otherwise, E (x, y) is 0;
step 3.2: if PN (b)k)∈{20,21,…,27,20+21,21+22,…,27+20Fifthly, turning to the step 3.3; otherwise, if NCz> α, mixing CzAdding the curve set C to a curve set C, and turning to the step 3.1; if N is presentCzNo more than α, curve C is not retainedzDirectly go to step 3.1, wherein NCzRepresents curve Czα is a predetermined minimum curve length;
step 3.3: if PN (b)k)∈{20,21,…,27Find out the coordinate in the set B satisfies | B }t.x-bk.x|+|bt.y-bkY | < 2 point btLet bk＝btDetermining bkIs curve CzA point above, let Cz＝Cz∪{bkAnd will point bkRemoving from the set B, and turning to the step 3.2; otherwise if PN (b)k)∈{20+21,21+22,…,27+20Find the coordinate in the set B satisfies | B }t.x-bk.x|+|bt.y-bkPoint b where y | ═ 1tLet bk＝btDetermining bkIs curve CzA point above, let Cz＝Cz∪{bkWill point bkRemoving from the set B, and turning to the step 3.2;
and 4, step 4: all curves C satisfying sigma larger than or equal to 0.4 in split curve set Ciσ is curve CiCalculating the straight line fitting error according to a formula (2), constructing a new curve set L according to all sub-curves obtained by splitting, and eliminating curves with the curve length smaller than α in the curve set L;
the splitting rule of the curve is as follows: constructing a straight line h according to two end points of the curve, finding a point on the curve which is farthest from the straight line h, taking the point as a splitting point to split the curve into two sub-curves, and continuously recursively splitting all the sub-curves which meet the condition that sigma is more than or equal to 0.4;
wherein r isjRepresents curve CiThe last j pixel point, NRIs curve CiThe number of upper points, h is the connection point r0And point
Straight line of (d)2(rjH) represents a point rjThe square of the vertical distance to the straight line h;
and 5: performing straight line fitting on each curve in the L by using a least square method to obtain a straight line set G, wherein each straight line G in the straight line set GiThe angle of (D) is denoted as Gi.θ，GiTheta is GiAn included angle with a horizontal straight line and G is less than or equal to 0 degreei.θ≤90°；
Step 6: determining the category number m of straight lines in the straight line set G, and specifically comprising the following steps:
step 6.1: creating a linear set O, and putting G in the linear set GiLine G with minimum thetaiAdding to the set O;
step 6.2: calculating each straight line G in the straight line set G according to the formula (3)iMinimum angular difference MinAngleD (G) from all the lines in Oi) If the minimum angle difference of the straight lines in the G is larger than β, the straight lines with the maximum minimum angle difference in the straight line set G are selected and added into the straight line set O, step 6.2 is carried out, wherein β is the preset maximum angle difference of similar straight lines, otherwise, if the minimum angle differences of the straight lines in the G are all smaller than β, m is equal to NOTurning to step 7;
MinAngleD(Gi)＝min{|Gi.θ-Oj.θ|,j＝0,1,…,NO-1} (3)
wherein min { } represents taking the minimum value in the set, OjTheta denotes the angle of the jth line in the set O, NORepresents the number of lines in the set O;
and 7: clustering all straight lines in G into m straight line subsets T by using K-means algorithmi，i＝0,1,…,m-1；
And 8: calculate each set of lines TiCorresponding straight line length sum, i is 0,1, …, m-1, finding the straight line set with the maximum straight line length sum, and marking as TkThen T iskThe straight line with the longest middle length is the key straight line of the image.
The invention has the advantages that: the image key straight line detection method comprises three parts of image preprocessing, straight line detection and key straight line judgment. The method utilizes the splitting method and the least square method to carry out the line detection, the algorithm efficiency and the accuracy are high, meanwhile, the K-means algorithm is utilized to carry out the line clustering firstly, and then the key line of the image is found according to the total length of the category line and the length of the line, so that the accuracy of the line detection is effectively improved, and the method provides guarantee for the processes of further target positioning, analysis, identification and the like.
Drawings
FIG. 1: a window corner image D;
FIG. 2: a Gaussian filtered window angle image;
FIG. 3: window corner edge detection map D1；
FIG. 4: according to the curve set C obtained in the step 3;
FIG. 5: according to the curve set L obtained in the step 4;
FIG. 6: according to the fitting straight line set G obtained in the step 5;
FIG. 7: obtaining a fitting straight line set G after screening;
FIG. 8: the image key line is determined according to steps 6-8.
Detailed Description
The following describes the process of the present invention in detail with reference to the specific vehicle window angle image example.
The invention discloses a method for detecting key straight lines of an image, which comprises the following steps:
step 1: zooming the original image into an image D with width and height respectively;
zooming the obtained window corner image according to the step 1 to obtain an image D as shown in fig. 1, wherein the width of the image D is 60px, and the height of the image D is 60 px;
step 2: carrying out Gaussian filtering and Canny edge detection operation on the image D to obtain an edge image D1An image D1Is B ═ Bi|i＝0,1,…,NB-1}，biRepresenting an image D1At an edge point of, and point biThe abscissa and ordinate of (A) are denoted by bi.x,bi.y，NBRepresenting the number of elements in set B;
performing Gaussian filtering on the image D according to the step 2 to obtain a filtered gray scale image as shown in FIG. 2, and further performing edge detection on the gray scale image by using a Canny operator to obtain an edge image D1As shown in fig. 3;
and step 3: obtaining an image D1The curve set C in (1) is specifically shown as steps 3.1-3.3;
step 3.1: if there is a sufficient PN (B) in set Bk)∈{20,21,…,27,20+21,21+22,…,27+20Point b ofkThen b iskIs determined as a new curve CzStarting point of (1), initializing curve Cz＝{bkWill point bkRemoving from the set B, and turning to the step 3.2; otherwise it states that all images D have been found1All the curves are added, and the step 4 is switched; wherein PN (b)k) Represents point bkThe eight neighborhood values are obtained by calculation according to a formula (1);
wherein, E (x, y) represents whether a point with coordinates (x, y) exists in the set B, if the point exists, E (x, y) is 1, otherwise, E (x, y) is 0;
step 3.2: if PN (b)k)∈{20,21,…,27,20+21,21+22,…,27+20B is describedkNot of curve CzTurning to step 3.3; otherwise, if NCz> α, curve C is determinedzAdding the curve to a curve set C to meet the length requirement, and turning to step 3.1; if N is presentCzNo more than α, curve C is not retainedzDirectly turning to the step 3.1; wherein N isCzRepresents curve Czα is a predetermined minimum curve length;
step 3.3: if PN (b)k)∈{20,21,…,27Find out the coordinate in the set B satisfies | B }t.x-bk.x|+|bt.y-bkY | < 2 point btLet bk＝btDetermining bkIs curve CzA point above, let Cz＝Cz∪{bkAnd will point bkRemoving from the set B, and turning to the step 3.2; otherwise if PN (b)k)∈{20+21,21+22,…,27+20Find the coordinate in the set B satisfies | B }t.x-bk.x|+|bt.y-bkPoint b where y | ═ 1tLet bk＝btDetermining bkIs curve CzA point above, let Cz＝Cz∪{bkWill point bkRemoving from the set B, and turning to the step 3.2;
according to step 3, image D1Curve extraction and screening are performed, and the obtained curve set C is shown in fig. 4, white dots on the graph represent end points of the curve, gray dots represent points on the curve, and the preset minimum curve length α is 10 in the example;
and 4, step 4: all curves C satisfying sigma larger than or equal to 0.4 in split curve set Ciσ is curve CiCalculating the straight line fitting error according to a formula (2), constructing a new curve set L according to all sub-curves obtained by splitting, and eliminating curves with the curve length smaller than α in the curve set L;
the splitting rule of the curve is as follows: constructing a straight line h according to two end points of the curve, finding a point on the curve which is farthest from the straight line h, taking the point as a splitting point to split the curve into two sub-curves, and continuously recursively splitting all the sub-curves which meet the condition that sigma is more than or equal to 0.4;
wherein r isjRepresents curve CiThe last j pixel point, NRIs curve CiThe number of upper points, h is the connection point r0And point
Straight line of (d)2(rjH) represents a point rjThe square of the vertical distance to the straight line h;
splitting all curves in the curve set C according to step 4 to obtain a new sub-curve set L as shown in fig. 5.
And 5: performing straight line fitting on each curve in the L by using a least square method to obtain a straight line set G, wherein each straight line G in the straight line set GiThe angle of (D) is denoted as Gi.θ，GiTheta is GiClip with horizontal straight lineAngle, and G is not less than 0 degreei.θ≤90°；
Performing straight line fitting by using a least square method according to the step 5 to obtain a fitted straight line set G as shown in FIG. 6, because a right bevel edge is expected in the present example, straight lines are further screened, and straight lines with straight line angles in the range of [0 degrees, 15 degrees ] in the straight line set G are removed, and the straight line screening result is shown in FIG. 7;
step 6: determining the category number m of straight lines in the straight line set G, and specifically comprising the following steps:
step 6.1: creating a linear set O, and putting G in the linear set GiLine G with minimum thetaiAdding to the set O;
step 6.2: calculating each straight line G in the straight line set G according to the formula (3)iMinimum angular difference MinAngleD (G) from all the lines in Oi) If the minimum angle difference of the straight lines in the G is larger than β, the straight lines with the maximum minimum angle difference in the straight line set G are selected and added into the straight line set O, step 6.2 is carried out, wherein β is the preset maximum angle difference of similar straight lines, otherwise, if the minimum angle differences of the straight lines in the G are all smaller than β, m is equal to NOTurning to step 7;
MinAngleD(Gi)＝min{|Gi.θ-Oj.θ|,j＝0,1,…,NO-1} (3)
wherein min { } represents taking the minimum value in the set, OjTheta denotes the angle of the jth line in the set O, NORepresents the number of lines in the set O;
the number m of straight line classes calculated according to step 6 is 3, and in this example the preset maximum angular difference β for similar straight lines is 4 °.
And 7: clustering all straight lines in G into m straight line subsets T by using K-means algorithmi，i＝0,1,…,m-1；
Clustering the detected 4 straight lines into 3 classes according to the step 7, and calculating that the angles of 6-9 of the 4 straight lines in the graph 7 are 74.95 degrees, 73.97 degrees, 70.04 degrees and 17.21 degrees respectively, so that the clustering results are 6 and 7, 8 and 9;
and 8: calculate each set of lines TiCorresponding straight lineThe length sum, i is 0,1, …, m-1, and a straight line set with the maximum straight line length sum is found and is marked as TkThen T iskThe straight line with the longest middle length is the key straight line of the image.
The key straight line of the image obtained according to step 8 is shown in fig. 8, as the straight line 6 in the figure.
The processing objects listed in the embodiments of the present invention are only used to illustrate the implementation process of the present invention, and the processing objects of the present invention are not limited to the illustrated examples.
Claims (2)
1. A method for detecting key straight lines of an image comprises the following steps:
step 1: zooming the original image into an image D with width and height respectively;
step 2: carrying out Gaussian filtering and Canny edge detection operation on the image D to obtain an edge image D1An image D1Is B ═ Bi|i＝0,1,…,NB-1}，biRepresenting an image D1At an edge point of, and point biThe abscissa and ordinate of (A) are denoted by bi.x,bi.y，NBRepresenting the number of elements in set B;
and step 3: obtaining an image D1Curve set C in (1);
and 4, step 4: all curves C satisfying sigma larger than or equal to 0.4 in split curve set Ciσ is curve CiCalculating the straight line fitting error according to a formula (2), constructing a new curve set L according to all sub-curves obtained by splitting, and eliminating curves with the curve length smaller than α in the curve set L;
the splitting rule of the curve is as follows: constructing a straight line h according to two end points of the curve, finding a point on the curve which is farthest from the straight line h, taking the point as a splitting point to split the curve into two sub-curves, and continuously recursively splitting all the sub-curves which meet the condition that sigma is more than or equal to 0.4;
wherein r isjRepresents curve CiThe last j pixel point, NRIs curve CiThe number of upper points, h is the connection point r0And point
Straight line of (d)2(rjH) represents a point rjThe square of the vertical distance to the straight line h;
and 5: performing straight line fitting on each curve in the L by using a least square method to obtain a straight line set G, wherein each straight line G in the straight line set GiThe angle of (D) is denoted as Gi.θ，GiTheta is GiAn included angle with a horizontal straight line and G is less than or equal to 0 degreei.θ≤90°；
Step 6: determining the category number m of straight lines in the straight line set G;
and 7: clustering all straight lines in G into m straight line subsets T by using K-means algorithmi，i＝0,1,…,m-1；
And 8: calculate each set of lines TiCorresponding straight line length sum, i is 0,1, …, m-1, finding the straight line set with the maximum straight line length sum, and marking as TkThen T iskThe straight line with the longest middle length is the key straight line of the image;
the step 3: obtaining an image D1The curve set C comprises the following specific steps:
step 3.1: if there is a sufficient PN (B) in set Bk)∈{20,21,…,27,20+21,21+22,…,27+20Point b ofkThen b iskIs determined as a new curve CzStarting point of (1), initializing curve Cz＝{bkWill point bkRemoving from the set B, and turning to the step 3.2;
otherwise, turning to the step 4; wherein PN (b)k) Represents point bkThe eight neighborhood values are obtained by calculation according to a formula (1);
wherein, E (x, y) represents whether a point with coordinates (x, y) exists in the set B, if so, E (x, y) is 1, otherwise, E (x, y) is 0;
step 3.2: if PN (b)k)∈{20,21,…,27,20+21,21+22,…,27+20Fifthly, turning to the step 3.3; otherwise, if NCz> α, mixing CzAdding the curve set C to a curve set C, and turning to the step 3.1; if N is presentCzNo more than α, curve C is not retainedzDirectly go to step 3.1, wherein NCzRepresents curve Czα is a predetermined minimum curve length;
step 3.3: if PN (b)k)∈{20,21,…,27Find out the coordinate in the set B satisfies | B }t.x-bk.x|+|bt.y-bkY | < 2 point btLet bk＝btDetermining bkIs curve CzA point above, let Cz＝Cz∪{bkAnd will point bkRemoving from the set B, and turning to the step 3.2; otherwise if PN (b)k)∈{20+21,21+22,…,27+20Find the coordinate in the set B satisfies | B }t.x-bk.x|+|bt.y-bkPoint b where y | ═ 1tLet bk＝btDetermining bkIs curve CzA point above, let Cz＝Cz∪{bkWill point bkAnd (5) removing the data from the set B, and turning to the step 3.2.
2. The image key straight line detection method according to claim 1, characterized in that: the step 6: determining the category number m of straight lines in the straight line set G, and specifically comprising the following steps:
step 6.1: creating a linear set O, and putting G in the linear set GiLine G with minimum thetaiAdding to the set O;
step 6.2: calculating each straight line G in the straight line set G according to the formula (3)iMinimum angular difference MinAngleD (G) from all the lines in Oi)，If the minimum angle difference of the straight lines in the G is larger than β, selecting the straight line with the maximum minimum angle difference in the straight line set G to be added into the straight line set O, and turning to step 6.2, wherein β is the preset maximum angle difference of similar straight lines, otherwise, if the minimum angle differences of the straight lines in the G are all smaller than β, making m equal to NOTurning to step 7;
MinAngleD(Gi)＝min{|Gi.θ-Oj.θ|,j＝0,1,…,NO-1} (3)
wherein min { } represents taking the minimum value in the set, OjTheta denotes the angle of the jth line in the set O, NORepresenting the number of lines in the set O.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710503951.4A CN107463939B (en) | 2017-06-27 | 2017-06-27 | Image key straight line detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710503951.4A CN107463939B (en) | 2017-06-27 | 2017-06-27 | Image key straight line detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107463939A CN107463939A (en) | 2017-12-12 |
| CN107463939B true CN107463939B (en) | 2020-06-09 |
Family
ID=60546112
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710503951.4A Expired - Fee Related CN107463939B (en) | 2017-06-27 | 2017-06-27 | Image key straight line detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107463939B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109035363B (en) * | 2018-07-05 | 2023-11-10 | 凌云光技术股份有限公司 | Line-circle optimal fitting method for rapid iteration |
| CN109255052B (en) * | 2018-08-29 | 2022-03-25 | 浙江工业大学 | Three-stage vehicle retrieval method based on multiple features |
| CN110874844A (en) | 2018-08-30 | 2020-03-10 | 京东方科技集团股份有限公司 | Line segment detection method, device and equipment |
| CN113705524A (en) * | 2021-09-06 | 2021-11-26 | 上海景吾智能科技有限公司 | Method and system for detecting vertical line in image |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012159869A (en) * | 2011-01-28 | 2012-08-23 | Daihatsu Motor Co Ltd | Image processing device |
| CN104949981A (en) * | 2014-03-24 | 2015-09-30 | 华中科技大学 | Automatic detection method and system for cotton five-euphylla period |
| CN106097329A (en) * | 2016-06-07 | 2016-11-09 | 浙江工业大学 | A kind of container profile localization method based on rim detection |
| CN106568424A (en) * | 2016-10-27 | 2017-04-19 | 西安科技大学 | Machine vision-based coal cutter rocker arm mining height detection method |
-
2017
- 2017-06-27 CN CN201710503951.4A patent/CN107463939B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012159869A (en) * | 2011-01-28 | 2012-08-23 | Daihatsu Motor Co Ltd | Image processing device |
| CN104949981A (en) * | 2014-03-24 | 2015-09-30 | 华中科技大学 | Automatic detection method and system for cotton five-euphylla period |
| CN106097329A (en) * | 2016-06-07 | 2016-11-09 | 浙江工业大学 | A kind of container profile localization method based on rim detection |
| CN106568424A (en) * | 2016-10-27 | 2017-04-19 | 西安科技大学 | Machine vision-based coal cutter rocker arm mining height detection method |
Non-Patent Citations (2)
| Title |
|---|
| 基于普通相机的交通事故现场三维重建关键技术研究;鲁光泉;《中国博士学位论文全文数据库 工程科技Ⅱ辑》;20041215;第C034-26页 * |
| 玻璃完整性在线检测关键算法研究及系统实现;胡旸;《中国优秀硕士学位论文全文数据库 工程科Ⅰ辑》;20130715;第B015-50页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107463939A (en) | 2017-12-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109145915B (en) | Rapid distortion correction method for license plate under complex scene | |
| CN107463939B (en) | Image key straight line detection method | |
| CN108052904B (en) | Method and device for acquiring lane line | |
| CN106934795A (en) | The automatic testing method and Forecasting Methodology of a kind of glue into concrete beam cracks | |
| CN101339601B (en) | License plate Chinese character recognition method based on SIFT algorithm | |
| CN109242870A (en) | A kind of sea horizon detection method divided based on image with textural characteristics | |
| CN111598952B (en) | Multi-scale cooperative target design and online detection identification method and system | |
| CN104036523A (en) | Improved mean shift target tracking method based on surf features | |
| CN103514448A (en) | Method and system for navicular identification | |
| CN105989334B (en) | Road detection method based on monocular vision | |
| CN103136525B (en) | High-precision positioning method for special-shaped extended target by utilizing generalized Hough transformation | |
| CN111444778A (en) | Lane line detection method | |
| CN108509950B (en) | Railway contact net support number plate detection and identification method based on probability feature weighted fusion | |
| CN109685827B (en) | Target detection and tracking method based on DSP | |
| CN114596500A (en) | Remote sensing image semantic segmentation method based on channel-space attention and DeeplabV3plus | |
| CN107862319B (en) | Heterogeneous high-light optical image matching error eliminating method based on neighborhood voting | |
| CN107564006B (en) | Circular target detection method utilizing Hough transformation | |
| CN108229247A (en) | A kind of mobile vehicle detection method | |
| CN113538503A (en) | Solar panel defect detection method based on infrared image | |
| CN113393426A (en) | Method for detecting surface defects of rolled steel plate | |
| CN115830359A (en) | Workpiece identification and counting method based on target detection and template matching in complex scene | |
| CN112861870A (en) | Pointer instrument image correction method, system and storage medium | |
| CN113989308B (en) | Polygonal target segmentation method based on Hough transformation and template matching | |
| CN113705564A (en) | Pointer type instrument identification reading method | |
| CN111626236A (en) | Rapid ellipse target detection method |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200609 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |