CN109215016B - Identification and positioning method for coding mark - Google Patents
Identification and positioning method for coding mark Download PDFInfo
- Publication number
- CN109215016B CN109215016B CN201810875464.5A CN201810875464A CN109215016B CN 109215016 B CN109215016 B CN 109215016B CN 201810875464 A CN201810875464 A CN 201810875464A CN 109215016 B CN109215016 B CN 109215016B
- Authority
- CN
- China
- Prior art keywords
- mark
- coding
- point
- image
- points
- 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
Images
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
-
- 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
- G06T9/00—Image coding
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for identifying and positioning a coding mark, which comprises the following steps: acquiring an image with a coding mark; preprocessing the acquired image to acquire a binary image; performing feature segmentation on the binary image, separating a circular region, performing feature template matching on a non-circular region, and judging whether the shape of a preset feature exists or not; acquiring a skeleton of a characteristic region in which a shape with preset characteristics is located; solving an affine transformation matrix according to the coordinates of the framework and specific parameters of preset characteristics; then carrying out inverse affine transformation, carrying out inverse transformation on the image coordinates of the dots, and converting the image coordinates into a coding mark design coordinate system; finding out dots corresponding to all the mark points matched with the design template on the image, and performing coordinate matching decoding on other dots to obtain coding points; and obtaining the unique code of the code mark according to the position corresponding to the code point. The identification positioning method improves the characteristic identification efficiency and the coding resolution precision aiming at the dynamic measurement requirement.
Description
Technical Field
The invention relates to the technical field of close-range photogrammetry, in particular to a method for identifying and positioning a coding mark.
Background
In the application field of close-range photogrammetry, the structure size of a measurement object is large, the curved surface is complicated, and the dynamism of the test object is a development trend. Due to the influence of factors such as camera resolution limitation, view field shielding, target offset caused by movement and the like, when a large-size structure in a large view field range is dynamically measured, the structure surface often lacks clear texture features with high identifiability, so that the feature information of the structure surface is inconvenient to directly, quickly and accurately extract, and the test requirement cannot be met. At present, the target point is usually generated by manually laying the cooperative mark on the structure to be detected for identification and tracking, so the design and application of the code mark with unique code value are widely researched and concerned.
Common coding marks are mainly classified into three types: point distribution, concentric rings, and a coded mark without distinct geometric features. The point distribution type coding mark is a digital code formed by introducing a design coordinate system according to the distribution of points on a plane, belongs to an absolute code and can generate a large number of codes, but a decoding and identifying algorithm is complex and time-consuming, for example, in the point distribution type coding mark technology, the method for identifying the point distribution type coding mark of the Chinese patent application with the application number of 201010201990.7 has the advantages that the positions of the designed coding points are closely arranged, and in order to ensure accurate decoding and identifying, the coding points in the same coding mark cannot be adjacent to each other, so that in a limited space of the designed coding points, the number of the arrangement and combination of the coding points is greatly limited, and the generation of a rich number of unique codes is difficult. Meanwhile, the coding mark needs at least 6 round points, when the coding mark is identified, more than 6 round points need to be extracted, each circle is judged, and therefore whether a mark point, a positioning point and a coding point exist is judged, whether a complete mark exists is confirmed, the calculation amount is large, and the decoding difficulty is large.
The concentric ring type coding mark is designed to be ring-shaped as a whole and consists of a central positioning mark and strip-shaped or point-shaped coding bits surrounding the central positioning mark. The type of coding mark has high positioning accuracy, but the coding quantity is limited, the coding belongs to relative coding, the initial coding bit needs to be judged, and the fuzzy of the image and the distortion of the annular coding region in dynamic measurement easily cause identification errors. In the concentric ring type coding mark technology, for example, chinese patent application publication No. CN105303224A discloses a coding mark point with large capacity and strong robustness and a decoding method thereof, the decoding method includes the following steps: s1: carrying out binarization processing on the obtained image, and extracting coding mark points; s2: extracting circular mark points according to the shape and the area of a white connected region in the coding mark points; s3: respectively calculating the distances from the circular mark points with larger area to other circular mark points and the circular mark points with larger area on the other side, judging the circular mark points with larger area and the distances from the circular mark points with larger area to all the circular mark points to be positioning mark points, judging the circular mark points with larger area on the other side to be initial mark points, and judging the other circular mark points to be coding points; s4: respectively calculating angles formed by two adjacent coding points from the initial mark point and the positioning mark point, and marking as alpha; and dividing alpha by m minus 1 to obtain the number of 0 of the interval between the two coding points, thereby obtaining the unique code corresponding to the coding mark point, wherein m is 360 DEG/k, and k is the number of coding bits. The decoding method improves the identification accuracy to a certain extent, but is mainly suitable for an annular coding mode, has low applicability to a point distribution type coding mode, and simultaneously, the characteristic identification efficiency and the coding resolving precision need to be improved when dynamic measurement and large-size structure splicing measurement are carried out.
Because the point distribution type and concentric ring type coding marks generally adopt the same-color circular marks, in the practical application process, because the inclination angle of the image is often larger, the camera has certain distortion, and the relation between the positions of the circular points in the coding marks is not stable enough. In order to solve the problem of identification of the coding mark, the chinese patent application No. 201310000907.3 discloses a method for decoding a point distribution type color coding mark, which comprises the steps of firstly identifying the color and shape of each graph in a photographic image, preliminarily determining the mark point of the color coding mark, identifying the mark point positioned in the center of the coding mark, finding out other 5 mark points by taking the mark point as the center, establishing an affine transformation model according to the set coordinates of 4 template points in the coding mark and the real coordinates thereof in the photographic image, calculating corresponding affine transformation parameters, and resolving the real coordinates of the coding point, thereby obtaining the code of the coding mark, simplifying the decoding step, improving the accuracy of identification of the color coding mark, but improving the hardware requirement of the method, when performing dynamic measurement and large-size structure splicing measurement, the problems of low feature recognition efficiency and low coding and resolving precision still exist.
The third type of coding marks without obvious features are generally formed by special shapes or combinations of special shapes, the coding quantity of the method is limited, and the image feature identification step is complicated.
Therefore, how to improve the feature recognition efficiency and the encoding calculation accuracy for the actual requirements of the encoding marks in the dynamic measurement and the large-size structure splicing measurement becomes a technical problem to be solved urgently by the technical personnel in the field.
Disclosure of Invention
In view of this, the invention provides a method for identifying and positioning a coding mark, which improves at least one of the efficiency of feature identification and the precision of code solution for the actual requirements of the coding mark in dynamic measurement and large-size structure splicing measurement.
The invention provides an identification and positioning method of a coding mark, wherein the coding mark comprises a mark point, a positioning point and a coding point which are positioned under a coding mark design coordinate system, the positioning point has a preset characteristic of distinguishing a dot, and the mark point and the coding point are dots, and the identification and positioning method comprises the following steps:
s20: acquiring an image with a coding mark;
s30: preprocessing the acquired image, and acquiring a binary image containing edge features;
s40: performing feature segmentation on the binary image, separating a circular region, performing feature template matching on a non-circular region, and judging whether the shape of a preset feature exists or not; if not, returning to the step S30;
s50: acquiring a skeleton of a characteristic region with a preset characteristic shape;
s60: solving an affine transformation matrix from the image coordinate system to the coding mark design coordinate system according to the coordinates of the framework and the design coordinates of the preset characteristics under the coding mark design coordinate system;
s70: searching for dots in a certain neighborhood range of preset characteristics, identifying the dots, solving the central coordinates of the dots, carrying out inverse affine transformation, carrying out inverse transformation on the image coordinates of the dots, and converting the image coordinates of the dots to a coding mark design coordinate system;
s80: finding out dots corresponding to all the mark points matched with the design template on the image, and carrying out coordinate matching decoding on other dots to obtain the coding values of the coding points;
s90: and obtaining the unique code of the code mark according to the position corresponding to the code point.
In the above method for identifying and locating a coded mark, step S50 preferably includes:
s51: acquiring a single-pixel framework of the characteristic region;
s52: and solving the normal direction of each point on the single-pixel framework, and calculating the optical center of the edge of the preset feature in the normal direction to obtain the sub-pixel framework of the edge.
In the above method for identifying and locating the coded mark, preferably, in step S51,
and acquiring a single-pixel skeleton of the characteristic region by adopting a K3M sequential iteration method.
In the above method for identifying and locating the coding mark, preferably, in step S30, the preprocessing includes one or more of filtering, denoising, and enhancing.
In the above method for identifying and locating the code mark, it is preferable that the circular region is separated by a shape parameter method in step S40.
Preferably, in step S80, finding out dots corresponding to all the mark points on the image that match the design template, and determining the correctness of the affine matrix solution, if not, returning to step S30; and then, performing coordinate matching decoding on other dots by a table look-up method to obtain the encoding value of the encoding point.
Preferably, in step S90, the method for identifying and locating the code mark further includes recording coordinates of the locating point and corresponding values of the code.
Preferably, the above method for identifying and locating a coded mark further includes, before step S20, step S10: and adhering a coding mark on the surface of the object to be measured, and acquiring an image.
Preferably, the mark points include a first mark point, a second mark point, and a third mark point, the mark points are intersections of a plurality of line segments, the code points include n points, n is an integer greater than or equal to 1, the first mark point, the second mark point, the third mark point, and the intersections are respectively located at four vertex angles of a square, the first mark point and the intersections are located on a diagonal of the square, an origin of a code mark design coordinate system is defined as the intersections, an X axis and a Y axis of the code mark design coordinate system are parallel to two adjacent sides of the rectangle, and the code points are distributed in the code mark design coordinate system in a staggered manner.
In the above method for identifying and locating a coded mark, preferably, the preset feature includes a cross shape.
The identification and positioning method of the coding mark provided by the invention aims at the actual requirements of the coding mark in dynamic measurement and large-size structure splicing measurement, improves the characteristic identification efficiency and the coding calculation precision through the steps S10-S90, and simultaneously provides a sub-pixel framework extraction method based on the thought of edge smoothness and minimum curvature change in the preferred scheme, so that the accurate mark positioning can be carried out. Meanwhile, the encoding mark is decoded based on the affine transformation theory and the table look-up method matching, the mark point method is simple in encoding, easy to identify, high in feature positioning precision and fast and stable in decoding algorithm.
Drawings
The drawings described herein are for illustration purposes only and are not intended to limit the scope of the present disclosure. In the drawings:
fig. 1 is a flowchart of a method for identifying and positioning a coded mark according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a design of a coding flag according to an embodiment of the present invention;
fig. 3 shows a design scheme of selecting 3 code points for the code flag in fig. 2.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
It should be noted that "first", "second", and "third" appearing in the present invention are only used for distinguishing the mark points, and are not distinguished in sequence, and are not used as the limitation of the mark points themselves; meanwhile, the term "and/or" in the present invention means that one of the schemes a and/or B may be provided or selected, for example, the three cases of the scheme a, the scheme B, the scheme a and the scheme B are included.
As shown in fig. 1 to 3, the present invention provides a method for identifying and positioning a coded mark, where the coded mark includes a mark point, a positioning point and a coding point located under a design coordinate system of the coded mark, the positioning point has a preset feature for distinguishing a dot, such as an intersection of a plurality of line segments, e.g., a midpoint of a square, a midpoint of a diamond, etc., as long as the positioning point has a typical preset feature, and the mark point and the coding point are dots, and the method for identifying and positioning includes the following steps:
s20: acquiring an image with a coding mark;
in this step, the acquired image may be an existing image or a newly acquired image, such as: before step S20, step S10 is further included: and adhering the coding mark on the surface of the object to be measured, and acquiring an image to obtain the image with the coding mark.
S30: preprocessing the acquired image, and acquiring a binary image containing edge features;
in this step, the preprocessing is mainly to improve the quality of the image, specifically, the preprocessing may include one or more of filtering, denoising, and enhancing, the denoising may be implemented in various ways, such as gaussian smoothing filtering, to obtain a smooth image, of course, the filtering and denoising may also be implemented by using a self-adaptive wiener filtering method, and the filtering and enhancing may also be implemented by using the existing way. Meanwhile, there are various ways to obtain the binary image, such as a bimodal method, a P parameter method, an iterative method, an OTSU method, and the like, and reference may be made to the prior art.
S40: performing feature segmentation on the binary image, separating a circular region, performing feature template matching on a non-circular region, and judging whether the shape of a preset feature exists or not; if not, returning to the step S30;
the feature segmentation can adopt an existing mode, such as an Otsu double threshold method, and preferably, a circular region can be separated through a shape parameter method, so that a non-circular region is obtained, and whether a shape with preset features exists in the non-circular region is conveniently analyzed, so that accurate and rapid positioning is performed. The characteristic template is a template with preset characteristics, whether the shape of the preset characteristics exists can be judged by matching the characteristic template with the segmented noncircular area, meanwhile, the quick search of the coding mark can be realized by matching the characteristic template with the preset characteristics due to the fact that the preset characteristics are different from the mark points (round points), and high-precision positioning can be realized by solving the coordinates of the preset characteristics.
Specifically, the principle of the shape parameter method is as follows: shape parameters include area, perimeter, circularity, rectangularity, curvature, and the like. Multiple parameters can be selected for screening according to needs in different object identifications. For example, using area A simultaneouslysEdge pixel perimeter LsCircularity ρcFor judging conditions, when the area whose value simultaneously meets the threshold is an ideal circular mark, and the rest are non-circular mark areas, thereby realizing the separation of the characteristic target and the background, and further searching and matching the non-circular area by using a template with preset characteristics (such as cross morphology). The formula is as follows:
s50: acquiring a skeleton of a characteristic region with a preset characteristic shape;
in this step, a skeleton of the feature region may be obtained in an existing manner, where the skeleton may be a pixel skeleton or other type of skeleton, the obtaining method may be Medial Axis Transform (MAT), iterative replacement algorithm, and the like, and specifically, with reference to the existing technique, the pixel skeleton may also include multiple forms, such as a single-pixel skeleton, a sub-pixel skeleton, and the like, and preferably, step S50 includes:
s51: acquiring a single-pixel framework of the characteristic region;
s52: and solving the normal direction of each point on the single-pixel framework, and calculating the optical center of the edge of the preset feature in the normal direction to obtain the sub-pixel framework of the edge.
Meanwhile, preferably, in step S51,
and acquiring a single-pixel skeleton of the characteristic region by adopting a K3M sequential iteration method.
The sub-pixel skeleton can be obtained by the following steps:
1) for contour point p on pixel skeletoni(x, y) (f (x, y) is the gray value of the pixel point), and the gradient vector g [ f ] is obtained according to the sobel operatorx,fy]；
2) Point piThe angle of the normal direction at (x, y) is
The direction of the notation line is (n)x,ny)；
3) After obtaining the normal direction, by pi(x, y) is used as a base point, a neighborhood range of one pixel is selected along the normal direction, 10 times of magnification interpolation is carried out on the neighborhood range, and a series of p 'can be obtained'i(x ', y '), wherein x ' ═ x + l · nx，y'＝y+l·ny，l∈(-0.5,0.5)；
4) At point p 'to be solved'iThe left neighborhood of (A) takes three known skeleton points pi-3,pi-2,pi-1Obtaining corresponding curvatures at every three adjacent points to obtain t1 and t2, and based on the idea of linear edge smoothing, making the curvature change | t2-t1| < epsilon, then p'i(x ', y') is the newly found skeleton point Pi；
5) Connecting all PiThus obtaining the sub-pixel framework.
For example, for the case that the preset feature of the positioning point is the intersection of a plurality of line segments, the end point coordinates of the line segment skeleton are read in sequence, and the coordinate values of the four end points are summed and averaged, so that the accurate positioning coordinates of the positioning point can be obtained.
S60: solving an affine transformation matrix from an image coordinate system to a coding mark design coordinate system according to the coordinates of the pixel framework and the design coordinates of the preset characteristics under the coding mark design coordinate system;
the image coordinate system, that is, the coordinate system in which the captured or acquired image is located, the affine transformation basic principle is as follows: the coded mark is imaged by photography, and affine transformation such as rotation, translation, scale scaling, distortion and the like can be generated in the picture. The affine transformation definition of the two-dimensional plane can be expressed as follows
x′＝Ax+b,x∈R2 (3)
Where b ═ b (b1, b2)T∈R2Is a translation matrix and
is a non-singular real matrix, and rotation is only a special case of affine transformation.
The affine transformation matrix a is specifically defined as follows:
where k is used to represent the scaling of the image, θ is used to represent the rotation angle of the image, and a and b are used to represent the tilt of the image in the x, y directions, respectively, we can get various affine transformations between the images according to different parameter combinations. In photogrammetry, the coding mark is adhered to a measured object, and only a small plane area is formed relative to the whole object, so that two-dimensional plane affine transformation is met, and therefore transformation matrix parameters from an image coordinate system to a coding mark design coordinate system at any position can be obtained.
S70: searching dots in a certain neighborhood range of preset characteristics, identifying the dots by using methods such as an ellipse fitting method and the like, solving the central coordinates of the dots, carrying out inverse affine transformation, carrying out inverse transformation on the image coordinates of the dots, and converting the image coordinates of the dots into a coding mark design coordinate system;
through the steps, the coordinates of the mark points and the coding points can be converted into a coding mark design coordinate system, so that the next step of identification and positioning is facilitated.
S80: finding out dots corresponding to all the mark points matched with the design template on the image, and carrying out coordinate matching decoding on other dots to obtain the coding values of the coding points;
the design template is a template with a whole coding mark (comprising mark points, locating points, coding points and the position relation of each point), specifically, in the step, round points corresponding to all the mark points matched with the design template on the image are found out, the accuracy of affine matrix solving is judged, and if the round points are incorrect, the step is returned to the step S30; and then, performing coordinate matching decoding on other dots by a table look-up method to obtain the encoding points.
S90: and obtaining a code with a unique code mark according to the position corresponding to the code point, wherein the code can be a serial number combination of the code point, a serial number corresponding to the serial number combination, or other codes. Preferably, this step may further include recording coordinates of the anchor point and the encoded corresponding value. That is, the location and identification of the coded mark is achieved.
In the above method for identifying and positioning a coding mark, the coding mark to be identified and positioned may be selected to meet the above requirements, specifically, as shown in fig. 2, the mark points include a first mark point a, a second mark point B, and a third mark point C, the mark point is an intersection O of a plurality of line segments, the coding point includes n points, n is an integer greater than or equal to 1, the first mark point a, the second mark point B, the third mark point C, and the intersection O are respectively located at four vertex angles of a square, the first mark point a and the intersection O are located on a diagonal line of the square, the origin of the coding mark design coordinate system is defined as the intersection O, and the X axis and the Y axis of the coding mark design coordinate system are parallel to two adjacent sides of the rectangle, the coding points are distributed in a staggered manner in the coding mark design coordinate system, the coding points at different positions should have unique coding numbers, and the coding numbers corresponding to the coding points are combined into a unique coding value.
In the above method for identifying and locating the code mark, the predetermined feature preferably includes a cross shape, but may also be a cross shape, or the like.
The coding mark is adopted in the identification and positioning method, the positioning point is easy to identify by adopting the positioning point (the intersection of a plurality of line segments) with the preset characteristics, whether the positioning point exists can be quickly judged through template matching, and quick and accurate positioning is realized, so that whether the coding mark exists is known, dot judgment (dot judgment in the prior art is uniformly adopted), and the calculation amount is small. In theory, the coding function can be realized by three marking points and one coding point (the total number is more than or equal to 4), the design structure is simpler, the number of marking circles is reduced, the position relation between the points is judged in the decoding process, and the calculation time is greatly shortened; meanwhile, when the positioning point is calculated through the sub-pixel framework, the positioning precision can be improved; the line characteristics of the positioning points and the length of the selected line segment of the design template can be used for conveniently solving the affine transformation matrix, and a guarantee is provided for decoding. In addition, the positioning points are cross-shaped intersection points, the quick search of the coding marks can be realized by utilizing cross template matching, and the high-precision positioning can be realized by solving the coordinates of the cross-shaped intersection points. Meanwhile, the first mark point A, the second mark point B, the third mark point C and the cross point O are respectively positioned on four vertex angles of a square, the first mark point A and the cross point O are positioned on a diagonal line of a rectangle, and the coding mark has strong anti-interference capability on imaging distortion, so that the reliability of the coding method is ensured.
In the above coding mark, the number of the coding points may be n equal to 1,2,3 …, the design positions of the coding points may be arranged as needed, the smaller n is, the more sparse the coding points are arranged, the higher the recognition and decoding accuracy is, but the fewer the design positions are, the larger n is, the greater n is, the number of the codes can be enriched, preferably, n is greater than 2, the connecting line of at least two coding points is parallel to the X axis, or/and the connecting line of at least two coding points is parallel to the Y axis, it should be noted that the parallel here is not absolute parallel, and due to the error of making and pasting the coding mark, it only needs to be approximately parallel. Furthermore, each coding point is arranged in rows and columns in the coding mark design coordinate system. In fig. 1, 29 encoding points are illustrated, including 3 rows parallel to the X-axis and 3 columns parallel to the Y-axis, it being understood that this is only one of the arrangements, and that further arrangements may be derived, such as 18, 25, 36, 43 encoding points, etc., including 5 rows parallel to the X-axis and 4 columns parallel to the Y-axis, etc. They are not listed and illustrated here, but are within the design framework of the invention.
Specifically, reference is made to fig. 2-3. The positioning points are cross-shaped with obvious graphic characteristicsAnd the cross points, the mark points and the coding points select round points, and different coding point combinations correspond to unique codes. According to the design principle, the coding capacity N conforms to the combinatory rules and has the formula:
wherein n is the total number of the designed coding labels, and m is the number of the selected coding points (m is more than or equal to 1 and is an integer). When n is 29 and m is 3, the coding capacity reaches 3654, which can basically meet all the test requirements (fig. 3(a) - (c) are design schemes for selecting 3 coding points, and the combination is three coding marks of 01-15-23, 06-11-25 and 01-17-25, although more coding marks can be selected). The cross-shaped cross point O is the origin of a coding mark design coordinate system, the X axis and the Y axis of the coordinate system are respectively parallel to AB and AC, and then the corresponding appointed coordinates of each marking point and each coding point can be obtained according to a designed proportion scale.
In general, the identification and positioning method of the invention improves the characteristic identification efficiency and the coding calculation precision through the steps S10-S90 aiming at the actual requirements of the coding mark in the dynamic measurement and the large-size structure splicing measurement, and simultaneously, in the preferable scheme, the sub-pixel framework extraction method is provided based on the thought of edge smoothness and minimum curvature change, so that the accurate mark positioning can be carried out. Meanwhile, the encoding mark is decoded based on the affine transformation theory and the table look-up method matching, the mark point method is simple in encoding, easy to identify, high in feature positioning precision and fast and stable in decoding algorithm.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (9)
1. A method for identifying and positioning a coded mark is characterized in that the coded mark comprises a mark point, a positioning point and a coding point which are positioned under a coded mark design coordinate system, the positioning point has a preset characteristic of distinguishing a dot, and the mark point and the coding point are dots, and the method for identifying and positioning the coded mark comprises the following steps:
s20: acquiring an image with a coding mark;
s30: preprocessing the acquired image, and acquiring a binary image containing edge features;
s40: performing feature segmentation on the binary image, separating a circular region, performing feature template matching on a non-circular region, and judging whether the shape of a preset feature exists or not; if not, returning to the step S30;
s50: acquiring a skeleton of a characteristic region with a preset characteristic shape;
s60: solving an affine transformation matrix from the image coordinate system to the coding mark design coordinate system according to the coordinates of the framework and the design coordinates of the preset characteristics under the coding mark design coordinate system;
s70: searching for dots in a certain neighborhood range of preset characteristics, identifying the dots, solving the central coordinates of the dots, carrying out inverse affine transformation, carrying out inverse transformation on the image coordinates of the dots, and converting the image coordinates of the dots to a coding mark design coordinate system;
s80: finding out dots corresponding to all the mark points matched with the design template on the image, and carrying out coordinate matching decoding on other dots to obtain the coding values of the coding points;
s90: obtaining a unique code of the code mark according to the position corresponding to the code point;
wherein, step S50 includes:
s51: acquiring a single-pixel framework of the characteristic region;
s52: and solving the normal direction of each point on the single-pixel framework, and calculating the optical center of the edge of the preset feature in the normal direction to obtain the sub-pixel framework of the edge.
2. The method for identifying and locating a coding mark according to claim 1, wherein in step S51,
and acquiring a single-pixel skeleton of the characteristic region by adopting a K3M sequential iteration method.
3. The method for identifying and locating a coding mark according to claim 1, wherein in step S30, the preprocessing includes one or more of filtering, denoising and enhancing.
4. The method for identifying and locating a coded mark according to claim 1, wherein in step S40, the circular area is separated by a shape parameter method.
5. The method for identifying and locating coding marks according to claim 1, wherein in step S80, the dots corresponding to all the mark points on the image matching the design template are found out, and the correctness of the affine matrix solution is determined, if not, the method returns to step S30; and then, performing coordinate matching decoding on other dots by a table look-up method to obtain the encoding value of the encoding point.
6. The method for identifying and locating a coded mark according to claim 1, further comprising recording coordinates of the located point and corresponding values of the codes in step S90.
7. The method for identifying and locating a coded mark according to any one of claims 1-6, further comprising, before step S20, step S10: and adhering a coding mark on the surface of the object to be measured, and acquiring an image.
8. The method for identifying and locating a coded mark according to any one of claims 1 to 6, it is characterized in that the mark points comprise a first mark point (A), a second mark point (B) and a third mark point (C), the anchor point is the intersection (O) of a plurality of line segments, the coding point comprises n points, n is an integer greater than or equal to 1, the first mark point (A), the second mark point (B), the third mark point (C) and the cross point (O) are respectively positioned on four vertex angles of a square, the first mark point (A) and the cross point (O) are positioned on a diagonal line of the square, the origin of a coding mark design coordinate system is defined as the cross point (O), and the X axis and the Y axis of the coding mark design coordinate system are parallel to two adjacent sides of the rectangle, and the coding points are distributed in the coding mark design coordinate system in a staggered manner.
9. The method for identifying and locating a coded mark according to any one of claims 1 to 6, wherein the predetermined features comprise a cross shape.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810875464.5A CN109215016B (en) | 2018-08-03 | 2018-08-03 | Identification and positioning method for coding mark |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810875464.5A CN109215016B (en) | 2018-08-03 | 2018-08-03 | Identification and positioning method for coding mark |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109215016A CN109215016A (en) | 2019-01-15 |
| CN109215016B true CN109215016B (en) | 2021-02-05 |
Family
ID=64987951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810875464.5A Active CN109215016B (en) | 2018-08-03 | 2018-08-03 | Identification and positioning method for coding mark |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109215016B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110163916B (en) * | 2019-05-23 | 2022-06-28 | 西华大学 | Rolling target-based surrounding type circular mark point matching method |
| CN111640083B (en) * | 2020-06-09 | 2022-07-08 | 中国科学院长春光学精密机械与物理研究所 | Color restoration method and system for object surface point cloud |
| CN113129394B (en) * | 2020-12-23 | 2022-09-06 | 合肥工业大学 | Parallelogram coding mark based on region division coding and coding method thereof |
| CN113188662B (en) * | 2021-03-16 | 2023-05-23 | 云南电网有限责任公司玉溪供电局 | Automatic infrared thermal imaging fault identification system and method |
| CN113313628B (en) * | 2021-06-29 | 2023-02-03 | 河南工业大学 | Affine transformation and mean pixel method-based annular coding point robustness identification method |
| CN113706621B (en) * | 2021-10-29 | 2022-02-22 | 上海景吾智能科技有限公司 | Mark point positioning and posture obtaining method and system based on marked image |
| CN116687569B (en) * | 2023-07-28 | 2023-10-03 | 深圳卡尔文科技有限公司 | Coded identification operation navigation method, system and storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7791620B2 (en) * | 2005-06-07 | 2010-09-07 | Ids Scheer Aktiengesellschaft | Systems and methods for rendering symbols using non-linear scaling |
| CN103310215A (en) * | 2013-07-03 | 2013-09-18 | 天津工业大学 | Detecting and identifying method for annular coding mark point |
| CN103530889A (en) * | 2013-11-04 | 2014-01-22 | 天津工业大学 | Method for decoding coding mark points of measuring stick based on skeleton extraction |
| CN104154877B (en) * | 2014-09-03 | 2016-07-06 | 中国人民解放军国防科学技术大学 | A kind of complicated convex-surface type object dimensional is rebuild and volume measuring method |
| US20180025479A1 (en) * | 2016-07-20 | 2018-01-25 | Toyota Motor Engineering & Manufacturing North America, Inc. | Systems and methods for aligning measurement data to reference data |
-
2018
- 2018-08-03 CN CN201810875464.5A patent/CN109215016B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109215016A (en) | 2019-01-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109215016B (en) | Identification and positioning method for coding mark | |
| CN108573511B (en) | Point-distributed cooperative coding mark and identification and positioning method thereof | |
| EP1232480B1 (en) | Image processing apparatus | |
| US7398928B2 (en) | Coded target and photogrammetry method using such targets | |
| CN110096920B (en) | High-precision high-speed positioning label and positioning method for visual servo | |
| CN107609451A (en) | A kind of high-precision vision localization method and system based on Quick Response Code | |
| CN107301368B (en) | Recognition method of DataMatrix two-dimensional code | |
| CN109190742B (en) | Decoding method of coding feature points based on gray feature | |
| CN112132907B (en) | Camera calibration method and device, electronic equipment and storage medium | |
| CN113712665B (en) | Positioning method and device based on positioning marker and computer storage medium | |
| JP6188052B2 (en) | Information system and server | |
| CN110415304B (en) | Vision calibration method and system | |
| CN115609591A (en) | 2D Marker-based visual positioning method and system and composite robot | |
| CN114792104A (en) | Method for identifying and decoding ring-shaped coding points | |
| CN113506276B (en) | Marker and method for measuring structural displacement | |
| CN113129397B (en) | Decoding method of parallelogram coding mark based on graphic geometric relation | |
| CN113793309B (en) | Subpixel level ellipse detection method based on morphological characteristics | |
| CN113313628B (en) | Affine transformation and mean pixel method-based annular coding point robustness identification method | |
| CN111256607B (en) | Deformation measurement method based on three-channel mark points | |
| CN111156917B (en) | Deformation measurement method based on gray level mark points | |
| CN113129396B (en) | Decoding method of parallelogram coding mark based on region segmentation | |
| CN113033233B (en) | Positioning method, device and storage medium for geometric deformation QR code | |
| CN115222828A (en) | Regular hexagon coding mark decoding method for visual positioning | |
| CN112750172A (en) | Improved method for point-shaped coding mark in photogrammetry | |
| Wang et al. | A novel cross-circular coded target for photogrammetry |
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 |