CN105405122A - Circle detection method based on data stationarity - Google Patents
Circle detection method based on data stationarity Download PDFInfo
- Publication number
- CN105405122A CN105405122A CN201510699508.XA CN201510699508A CN105405122A CN 105405122 A CN105405122 A CN 105405122A CN 201510699508 A CN201510699508 A CN 201510699508A CN 105405122 A CN105405122 A CN 105405122A
- Authority
- CN
- China
- Prior art keywords
- circle
- center
- image
- variance
- arc profile
- 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
Classifications
-
- 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
- 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
- G06T2207/30164—Workpiece; Machine component
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a circle detection method based on data stationarity. The method comprises the following steps of step1, through image preprocessing, acquiring an arc contour of a single pixel width; step2, using three different points on the arc contour of the single pixel width to determine a position of a theory circle center; step3, calculating distance data from each possible circle center to each pixel point on the arc contour of the single pixel width; step4, searching an actual circle center and an actual radius, wherein after the actual circle center is acquired, a mean value of the distance data corresponding to the actual circle is a radius of the circle. By using the method, an occupation memory is less and a calculated amount is less too; and real-time performance and algorithmic accuracy are high.
Description
Technical field
The invention belongs to graphical analysis and identification field, be specifically related to a kind of circle detection method based on data stationarity.
Technical background
Along with the development of artificial intelligence, its to the accuracy of image recognition and requirement of real-time more and more high, no matter be in military affairs, industry spot or daily life, a lot of object exists with circle, in order to realize the location of circular object, facilitate the crawl of mechanical arm, effectively and detect that round position and dimensional parameters seem very necessary in real time.
Common circle detection method has classical Hough transforms loop truss method and random Hough transformation loop truss method.Hough transform is a kind of technique of image edge detection that PaulHough proposed in 1962, and it can detect any analytic curve in image space.The outstanding advantages of Hough transform is the local peak detection global detection in image be converted in parameter space, and in addition, Hough transform has good robustness to random noise and can detect the incomplete analytic curve of profile, is thus widely used.
But in classical Hough transforms, image space is one to many conversion to the conversion of parameter space, and therefore committed memory is large, calculated amount large, and real-time is not high.
For the defect that classical Hough transforms exists, random Hough transformation arises at the historic moment, but it is due to aimless stochastic sampling, causes a large amount of invalid accumulative, reduces the accuracy of algorithm.
Summary of the invention
The technical problem to be solved in the present invention is existing circle detection method or committed memory is large, calculated amount is large, and real-time is not high; Or can cause a large amount of invalid accumulative, the accuracy of algorithm is low.
For solving the problems of the technologies described above, the technical solution used in the present invention is: a kind of circle detection method based on data stationarity, comprises the following steps: step 1: the arc profile being obtained single pixel wide degree by Image semantic classification; The arc profile of described single pixel wide degree refers at circle in the radial direction only have a pixel to belong to arc profile; Step 2: three different points on the arc profile getting arbitrarily single pixel wide degree, connect 3 between two, obtain three strings, appoint and get two strings, do its perpendicular bisector respectively, obtain the theoretical center of circle by the intersection point of two perpendicular bisectors to two strings; Step 3: centered by the described theoretical center of circle, the fluctuation range width δ in a given center of circle, the unit of δ is pixel, calculate the range data of each pixel on each possibility center of circle to the arc profile of described single pixel wide degree, in order to form distance matrix, obtain the variance often organizing range data; Described distance matrix is three-dimensional matrice, the first peacekeeping two-dimensional characterization be described may the coordinate in the center of circle, what the third dimension characterized is with described in certain may the center of circle for the center of circle time, the distance of each pixel on the arc profile of itself and described single pixel wide degree; Step 4: use the variance that range data corresponding to each possibility center of circle is tried to achieve, form variance matrix, described variance is the variance of distance matrix third dimension data under each coordinate, and described variance matrix is two-dimensional matrix; Step 5: the minimum value finding out described variance matrix, the possible center of circle corresponding to this minimum value is the true center of circle, and the average of the described range data that the described true center of circle is corresponding is radius.
The present invention is compared with classical Hough transforms loop truss method, and avoid one to many spatial alternations, real-time is higher; There is good robustness and the incomplete circular arc of profile can be detected; The present invention is compared with random Hough transformation loop truss, and any 3 Position Approximates determining the center of circle on the present invention's circular arc, this step finally determines that the impact of home position is less, so accuracy is higher to maintenance data stationarity in subsequent step.In addition, due to the impact that the factor such as noise, image processing process produces, deviation in various degree can be there is in the described theoretical center of circle, in order to overcome this problem, the fluctuation range width δ in a given center of circle, the unit of δ is pixel, then may the position range in the center of circle be, at once the column width matrix that is all; Described distance vector, refer to some may the center of circle for the center of circle time, the distance of each pixel on the described possibility center of circle to the arc profile of single pixel wide degree, this is a multi-C vector, and the dimension of vector equals the pixel number of the arc profile of single pixel wide degree; Described distance matrix is the combination of the described range data obtained by all described possibility centers of circle; Described variance matrix asks variance to distance vector described in each respectively, then the variance that all described possibility centers of circle obtain combined; From mathematics general knowledge, on circle, each point is equal to the distance in the center of circle, and in other words, the last point of circle is to the distance in the center of circle, and the variance of these group data is zero.Consider the impact of digital picture self character, the arc profile of the described single pixel wide degree of image procossing gained can not be zero to the variance of the distance in the center of circle, can only as far as possible close to zero.Therefore, find out the least member in described variance matrix, it correspond to position, the true center of circle, and the average of the described data vector of its correspondence is true radius.
Concrete, in described step 1, pretreated method comprises the following steps: 1) by pending image gray processing, obtains the gray level image of single pass data; 2) gray level image is carried out binary conversion treatment, obtain binary image; 3) Morphological scale-space is carried out to binary image and remove image noise; 4) skeletal extraction is carried out to the binary image removing image noise; 5) carry out deburring process to the image after skeletal extraction, obtain this kind of arc profile, make it belong to round at the pixel that only has in the radial direction of circle, rest of pixels point all belongs to background.
Advantage of the present invention is: committed memory is little, calculated amount is little, and real-time is high, the accuracy of algorithm is high.
Accompanying drawing explanation
Fig. 1 is the loop truss process flow diagram of the embodiment of the present invention.
Fig. 2 is the schematic diagram of circle detection method of the present invention.
Fig. 3 is that the image processing process of the embodiment of the present invention is shown.
Fig. 4 is that the loop truss result of the embodiment of the present invention is shown, (center of circle is marked by "+", and what dotted line "--" showed is the circle drawn according to the center of circle detected and radius).
Embodiment
Embodiment:
Can hold to clearly set forth technology of the present invention, being explained especially exemplified by example with reference.
Fig. 1 is process flow diagram of the present invention, and image, through Image semantic classification flow process, obtains the arc profile of single pixel wide degree; Again the arc profile of described single pixel wide degree is used for characteristic area to determine, particularly, determines the theoretical center of circle with any 3 on the arc profile of described single pixel wide degree, then centered by the theoretical center of circle, the fluctuation range in the given possibility center of circle; Then, characteristic quantity calculating may carried out, in particular to distance matrix and variance matrix within the scope of the center of circle; Finally, accurate location and the radius in the center of circle is determined by the characteristic of circle and the relevant knowledge of data stationarity.
Fig. 2 is the schematic diagram of circle detection method of the present invention, determines the theoretical center of circle by any POINT1, POINT2 and POINT3 on the arc profile of described single pixel wide degree at 3; Centered by the theoretical center of circle, the fluctuation range width δ in the given possibility center of circle, the unit of δ is pixel, then the position in the possibility center of circle is matrixes that ranks width is all; Ask every bit on this matrix, namely each may the distance D (i of each pixel on center of circle to the arc profile of described single pixel wide degree, j, k), finally obtain described distance matrix, described distance matrix is a three-dimensional matrice, the first peacekeeping two-dimensional characterization be described may the coordinate in the center of circle, what the third dimension characterized is with described in certain may the center of circle for the center of circle time, the distance of each pixel on the arc profile of itself and described single pixel wide degree.
Fig. 3 is pretreated method in step 1 of the present invention, comprises the following steps:
1) by pending image gray processing, the gray level image of single pass data is obtained;
2) gray level image is carried out binary conversion treatment, obtain binary image;
3) Morphological scale-space is carried out to binary image and remove image noise;
4) cutting is carried out to the binary image removing image noise, cut out the area image at Circle in Digital Images place;
5) skeletal extraction is carried out to the image that cutting obtains;
6) carry out deburring to skeleton image, obtain this kind of arc profile, make it belong to round at the pixel that only has in the radial direction of circle, rest of pixels point all belongs to background, the image that namely final pre-service is complete.
Fig. 4 is that the loop truss result of the embodiment of the present invention is shown, as can be seen from this figure, the circle detection method based on data stationarity that the present invention proposes has very high precision.
Below by reference to the accompanying drawings embodiments of the present invention are described in detail, but the present invention is not limited to described embodiment.For those of ordinary skill in the art, in the scope of principle of the present invention and technological thought, multiple change, amendment, replacement and distortion are carried out to these embodiments and still falls within the scope of protection of the present invention.
Claims (2)
1., based on a circle detection method for data stationarity, it is characterized in that comprising the following steps:
Step 1: the arc profile being obtained single pixel wide degree by Image semantic classification; The arc profile of described single pixel wide degree refers at circle in the radial direction only have a pixel to belong to arc profile;
Step 2: three different points on the arc profile getting arbitrarily single pixel wide degree, connect 3 between two, obtain three strings, appoint and get two strings, do its perpendicular bisector respectively, obtain the theoretical center of circle by the intersection point of two perpendicular bisectors to two strings;
Step 3: centered by the described theoretical center of circle, the fluctuation range width δ in a given center of circle, the unit of δ is pixel, calculate the range data of each pixel on each possibility center of circle to the arc profile of described single pixel wide degree, in order to form distance matrix, obtain the variance often organizing range data; Described distance matrix is three-dimensional matrice, the first peacekeeping two-dimensional characterization be described may the coordinate in the center of circle, what the third dimension characterized is with described in certain may the center of circle for the center of circle time, the distance of each pixel on the arc profile of itself and described single pixel wide degree;
Step 4: use the variance that range data corresponding to each possibility center of circle is tried to achieve, form variance matrix, described variance is the variance of distance matrix third dimension data under each coordinate, and described variance matrix is two-dimensional matrix;
Step 5: the minimum value finding out described variance matrix, the possible center of circle corresponding to this minimum value is the true center of circle, and the average of the described range data that the described true center of circle is corresponding is radius.
2. the circle detection method based on data stationarity according to claims 1, is characterized in that, in described step 1, pretreated method comprises the following steps:
1) by pending image gray processing, the gray level image of single pass data is obtained;
2) gray level image is carried out binary conversion treatment, obtain binary image;
3) Morphological scale-space is carried out to binary image and remove image noise;
4) skeletal extraction is carried out to the binary image removing image noise,
5) carry out deburring process to the image after skeletal extraction, obtain this kind of arc profile, make it belong to round at the pixel that only has in the radial direction of circle, rest of pixels point all belongs to background.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510699508.XA CN105405122A (en) | 2015-10-26 | 2015-10-26 | Circle detection method based on data stationarity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510699508.XA CN105405122A (en) | 2015-10-26 | 2015-10-26 | Circle detection method based on data stationarity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105405122A true CN105405122A (en) | 2016-03-16 |
Family
ID=55470589
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510699508.XA Pending CN105405122A (en) | 2015-10-26 | 2015-10-26 | Circle detection method based on data stationarity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105405122A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106204542A (en) * | 2016-06-29 | 2016-12-07 | 上海晨兴希姆通电子科技有限公司 | Visual identity method and system |
| CN107067432A (en) * | 2017-04-10 | 2017-08-18 | 武汉理工大学 | The determination method of round edge circle in quartz pushrod detection |
| CN107981791A (en) * | 2017-12-04 | 2018-05-04 | 深圳市沃特沃德股份有限公司 | Cleaning method, device and the vision sweeper of vision sweeper |
| CN109869109A (en) * | 2019-03-26 | 2019-06-11 | 中铁大桥局集团有限公司 | A kind of method and drill bit fishing device for salvaging drill bit |
| WO2019109227A1 (en) * | 2017-12-04 | 2019-06-13 | 深圳市沃特沃德股份有限公司 | Cleaning method and device for vision-aided robotic vacuum cleaner, and vision-aided robotic vacuum cleaner |
| CN110824451A (en) * | 2019-11-20 | 2020-02-21 | 上海眼控科技股份有限公司 | Processing method and device of radar echo map, computer equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556145A (en) * | 2009-04-29 | 2009-10-14 | 徐州工程学院 | Device and method for monitoring slow-speed and over-load rotor eccentricity image |
| JP2011134092A (en) * | 2009-12-24 | 2011-07-07 | Honda Motor Co Ltd | Image processing apparatus, image processing method, and image processing program |
| CN104764465A (en) * | 2015-03-30 | 2015-07-08 | 中国资源卫星应用中心 | Method for measuring ground resolution of remote sensing satellite |
-
2015
- 2015-10-26 CN CN201510699508.XA patent/CN105405122A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556145A (en) * | 2009-04-29 | 2009-10-14 | 徐州工程学院 | Device and method for monitoring slow-speed and over-load rotor eccentricity image |
| JP2011134092A (en) * | 2009-12-24 | 2011-07-07 | Honda Motor Co Ltd | Image processing apparatus, image processing method, and image processing program |
| CN104764465A (en) * | 2015-03-30 | 2015-07-08 | 中国资源卫星应用中心 | Method for measuring ground resolution of remote sensing satellite |
Non-Patent Citations (2)
| Title |
|---|
| 刘志奎: "基于计算机视觉的链条元件几何参数测量", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 张帅: "基于中点画圆的圆检测", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106204542A (en) * | 2016-06-29 | 2016-12-07 | 上海晨兴希姆通电子科技有限公司 | Visual identity method and system |
| CN106204542B (en) * | 2016-06-29 | 2019-04-19 | 上海晨兴希姆通电子科技有限公司 | Visual identity method and system |
| CN107067432A (en) * | 2017-04-10 | 2017-08-18 | 武汉理工大学 | The determination method of round edge circle in quartz pushrod detection |
| CN107981791A (en) * | 2017-12-04 | 2018-05-04 | 深圳市沃特沃德股份有限公司 | Cleaning method, device and the vision sweeper of vision sweeper |
| WO2019109227A1 (en) * | 2017-12-04 | 2019-06-13 | 深圳市沃特沃德股份有限公司 | Cleaning method and device for vision-aided robotic vacuum cleaner, and vision-aided robotic vacuum cleaner |
| CN107981791B (en) * | 2017-12-04 | 2020-01-14 | 深圳市无限动力发展有限公司 | Sweeping method and device of visual sweeper and visual sweeper |
| CN109869109A (en) * | 2019-03-26 | 2019-06-11 | 中铁大桥局集团有限公司 | A kind of method and drill bit fishing device for salvaging drill bit |
| CN110824451A (en) * | 2019-11-20 | 2020-02-21 | 上海眼控科技股份有限公司 | Processing method and device of radar echo map, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105405122A (en) | Circle detection method based on data stationarity | |
| CN108562274B (en) | Marker-based non-cooperative target pose measurement method | |
| CN104331876B (en) | Method for detecting straight line and processing image and related device | |
| CN109784250B (en) | Positioning method and device of automatic guide trolley | |
| CN111723721A (en) | Three-dimensional target detection method, system and device based on RGB-D | |
| CN110942515A (en) | Point cloud-based target object three-dimensional computer modeling method and target identification method | |
| CN110780276B (en) | Tray identification method and system based on laser radar and electronic equipment | |
| CN112232248B (en) | Method and device for extracting plane features of multi-line LiDAR point cloud data | |
| CN111598946A (en) | Object pose measuring method and device and storage medium | |
| CN112017232A (en) | Method, device and equipment for positioning circular pattern in image | |
| CN104331885A (en) | Circular target detection method based on voting line clustering | |
| Sansoni et al. | Optoranger: A 3D pattern matching method for bin picking applications | |
| CN116051822A (en) | Concave obstacle recognition method and device, processor and electronic equipment | |
| CN112861870A (en) | Pointer instrument image correction method, system and storage medium | |
| CN113298838B (en) | Object contour line extraction method and system | |
| US9443312B2 (en) | Line parametric object estimation | |
| CN116091727A (en) | Complex Qu Miandian cloud registration method based on multi-scale feature description, electronic equipment and storage medium | |
| Moritani et al. | Cylinder-based efficient and robust registration and model fitting of laser-scanned point clouds for as-built modeling of piping systems | |
| CN110673607A (en) | Feature point extraction method and device in dynamic scene and terminal equipment | |
| CN111986299A (en) | Point cloud data processing method, device, equipment and storage medium | |
| CN117745780A (en) | Outdoor large scene 3D point cloud registration method based on isolated cluster removal | |
| CN113744323B (en) | Point cloud data processing method and device | |
| CN112017290B (en) | Conical surface fitting method and device, computer equipment and storage medium | |
| CN108917768B (en) | Unmanned aerial vehicle positioning navigation method and system | |
| CN113627548A (en) | Planar workpiece template matching method, device, medium and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160316 |