CN109242957A - A kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint - Google Patents
A kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint Download PDFInfo
- Publication number
- CN109242957A CN109242957A CN201810980173.2A CN201810980173A CN109242957A CN 109242957 A CN109242957 A CN 109242957A CN 201810980173 A CN201810980173 A CN 201810980173A CN 109242957 A CN109242957 A CN 109242957A
- Authority
- CN
- China
- Prior art keywords
- matrix
- single frames
- arrays
- projector
- dimensional
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- 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
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The present invention relates to a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint, include the following steps, S1 projects single frames coded structured light to testee surface by projector, and the single frames coded structured light of the projector projects is tied on horizontal epipolar line;S2 shoots the coded image of the single frames coded structured light deformed by the testee surface modulation using camera;S3 is decoded the coded image using column direction constraint, horizontal epipolar line constraint and three kinds of the constraint relationships of coded sequence, obtains position of each encoded point in coded image in coded image;S4 according to the system parameter demarcated in advance, and combines position of each encoded point in coded image in coded image, calculates the three dimensional point cloud on the testee surface.The precision and resolution ratio of three-dimensional reconstruction can be improved in the present invention, reduce error rate.
Description
Technical field
The present invention relates to three-dimensional reconstruction fields, and in particular to a kind of single frames coding structural light three-dimensional weight based on multiple constraint
Construction method.
Background technique
It is different according to the form of institute's projection source in active vision Three Dimensional Reconfiguration, it is divided into point, line and area-structure light three
Kind mode, they are utilized respectively luminous point, light and two-dimensional pattern, so that tune of institute's projection source by testee space characteristics
System.In two-dimentional area-structure light mode, the structured light patterns containing encoded information are projected into tested scene frequently with projector
Or target surface, projector is equivalent to a reverse video camera at this time, it does not need to take multiple scan during three-dimensionalreconstruction,
A video camera or other optical sensors need to be only used, the piece image of tested scene or object is shot, passes through feature detection, solution
The means such as code, image procossing and data fitting can obtain testee surface three dimension information, and hardware realization is convenient and sampling is fast
Degree is fast.It mainly include time encoding and space encoding from the point of view of the coding mode of area-structure light pattern.Time encoding is will be different
Coding pattern chronologically successively project body surface, obtain corresponding coded video sequences, which be decoded, it is excellent
Point is that sampling density is high, accuracy of measurement is high, but need to acquire more coded image causes measuring speed low, therefore is not suitable for
It is applied in dynamic 3 D measurement.Space encoding (also referred to as single frames coding) is that one width of projector projects encodes in some way
Image, camera receive and parse through corresponding coded image, obtain the light on image on body surface each point corresponding to each picture point
Line projection angle, finally obtains the depth information of object by trigonometry principle, and this method speed is suitable for dynamic fastly and measures.Currently,
Structure light coding based on single frames has following methods:
(1) method for reconstructing of colored structures pumped FIR laser
Based on RGB true color coding, color fringe coding, colored Lattice encoding, colour dot lattice coding is according to pixel
Color-values define coding pattern code word, this method decoding difficulties, and accuracy heavy dependence projects consistent with acquisition color value
Property, very sensitive to noise and object color, three-dimensionalreconstruction error rate is big.
(2) based on the method for reconstructing of Fourier transformation
Its cardinal principle of Fourier transformation is the stripe pattern (such as Moire fringe) of the grating or different wave length to deformation
It is converted, after obtaining fundamental component, acquires phase value using inversefouriertransform.Fourier transformation need to only acquire a secondary change
The stripe pattern of shape can restore the 3D shape of entire testee, be very suitable for real-time dynamic 3 D measurement.But Fourier
There are computationally intensive, it is also easy to produce spectral aliasing, it is multiple along with being influenced to be unable to measure Fourier transformation by maximum measurement gradient
Miscellaneous object;Also, the reflection characteristic on Fourier transform pairs testee surface is very sensitive, has seriously affected measurement accuracy.
(3) method for reconstructing based on gray scale index
Pseudo-random sequence has window uniqueness, and each different subsequence can find absolute position in entire sequence
It sets, is widely used in space compression schemes.Typically, the two-dimensional expansion M-arrays of DeBrujin sequence usually combines item
Line, filament, grid, dot, simple geometric shape of designed, designed etc. carry out three-dimensional reconstruction, the party convenient for the basic element of identification
Method only needs to project a frame grating, meets the needs of dynamic measures, but its Measurement Resolution is relatively low, to the company of body surface
Continuous property has certain requirement.
Summary of the invention
Technical problem to be solved by the invention is to provide a kind of single frames coding structural light three-dimensional weight based on multiple constraint
The precision and resolution ratio of three-dimensional reconstruction can be improved in construction method, reduces error rate.
The technical scheme to solve the above technical problems is that a kind of single frames coded structured light based on multiple constraint
Three-dimensional rebuilding method includes the following steps,
S1 projects single frames coded structured light to testee surface by projector, and by the list of the projector projects
Frame coding structure light is tied on horizontal epipolar line;
S2 shoots the coding shadow of the single frames coded structured light deformed by the testee surface modulation using camera
Picture;
S3, using column direction constraint, horizontal epipolar line constraint and three kinds of the constraint relationships of coded sequence to the coded image into
Row decoding, obtains position of each encoded point in the coded image in the coded image;
S4, according to the system parameter demarcated in advance, and in conjunction with each encoded point in the coded image in the coded image
In position, calculate the three dimensional point cloud on the testee surface.
The beneficial effects of the present invention are: in a kind of single frames coding structural light three-dimensional reconstruction side based on multiple constraint of the present invention
Two-dimensional search in order to preferably extend and identify structure light coding image, and is reduced to one-dimensional linear search by method, and the present invention mentions
Structure light coding image is constrained on horizontal epipolar line out, is conducive to line direction and indexes, enormously simplify the complexity of problem and mention
High operation efficiency, reduces error rate;
The present invention uses coded sequence the constraint relationship and column direction the constraint relationship, and more letters can be provided for subsequent processing
Breath accelerates decoding efficiency, improves robustness and resolution ratio.
Based on the above technical solution, the present invention can also be improved as follows.
Further, in the S1, the single frames coded structured light of the projector projects is tied on horizontal epipolar line
Specific method is,
By putting the projector and the camera around the testee, make the camera and the projector
Between baseline be overlapped with the X-axis of the projector;
Wherein, the baseline between the camera and the projector is specifically, the installation center of the camera and the throwing
Line between the projection plane center of shadow instrument;The X-axis of the projector is specifically, with the projection plane center of the projector
Make XYZ coordinate system for origin, the X-axis in XYZ coordinate system is the X-axis of the projector.
Further, the coding method of the single frames coded structured light of the projector projects is using DeBrujin sequence
Two-dimensional expansion M-arrays matrix;
The two-dimensional expansion M-arrays matrix is made of the stacking of multiple M-arrays matrixes, and each M-arrays
Matrix is formed by DeBrujin sequence structure;
The child window matrix H of each of M-arrays matrix m*n is in the entire two-dimensional expansion M-arrays
There is uniqueness in matrix, and, N=kmn- 1=R*C, wherein N is the length of the DeBrujin sequence, and k is described
The number of element figure in DeBrujin sequence, R are the lateral resolution of the two-dimensional expansion M-arrays matrix, and C is described
The longitudinal resolution of two-dimensional expansion M-arrays matrix.
Further, the line number of the M-arrays matrix is 4, then the R=4.
Beneficial effect using above-mentioned further scheme is: the present invention uses improved DeBrujin sequence coding method,
It may be implemented also achieve the purpose that Measurement Resolution is high in the case where element figure number is few.
It further, include wide, narrow two kinds of element figures in the single frames coded structured light of the projector projects, with right respectively
It should indicate 1 in binary system and 0;
According to wide, narrow two kinds of element figures, and combine the coding of the two-dimensional expansion M-arrays matrix of DeBrujin sequence
Method obtains the comb mesh pattern that a kind of vertical narrow strips intersect composition with horizontal wide striped, and the comb mesh pattern is the throwing
The pattern that the single frames coded structured light that shadow instrument projects is formed.
Beneficial effect using above-mentioned further scheme is: the present invention uses wide, narrow two kinds of element figures, in narrow strips
Heart line ensure that the consistency of column direction, improve decoded stability and robustness;In addition, by column direction the constraint relationship,
The three-dimensional coordinate position of each pixel on available center line, by pass point cloud, compared to previous based on DeBrujin
The single frames reconstruction mode of sequence, Measurement Resolution is higher, and details is richer.
Further, the horizontal wide striped is specially horizontal circle angular width striped.
Beneficial effect using above-mentioned further scheme is: horizontal width striped is acquired using rounded shapes in camera
To coded image in oval form is presented, it is subsequent accurately to be extracted by sub-pixel edge positioning method combination ellipse fitting
Measurement accuracy is improved in the position of encoded point.
Further, the S3 specifically,
S31 extracts the encoded point in the coded image, is existed by column direction the constraint relationship and horizontal epipolar line the constraint relationship
The upper and lower, left and right connection relationship between the encoded point is established in the two-dimensional expansion M-arrays matrix, obtains multiple institutes
State child window matrix H;
The M-arrays matrix at each child window matrix H and place is carried out encoding array respectively by S32
Match, determines position of each encoded point in the M-arrays matrix in each child window matrix H;
S33 indexes out each encoded point in each child window matrix H using horizontal epipolar line the constraint relationship respectively and is encoding
Position in image.
Further, in the S31, by extracting the center line of vertical narrow strips, the column direction constraint of encoded point is obtained
Relationship.
Further, in the S32, position of each encoded point in the M-arrays matrix in the child window matrix H
It sets and respectively includes line number and row number;Line number phase of each encoded point in the M-arrays matrix in the child window matrix H
It is relative position for the entire two-dimensional expansion M-arrays matrix, each encoded point is in institute in the child window matrix H
Stating the row number in M-arrays matrix is absolute position for the entire two-dimensional expansion M-arrays matrix;
In the S33, is indexed out in each child window matrix H using horizontal epipolar line the constraint relationship and respectively encoded respectively
Absolute line number of the point in the two-dimensional expansion M-arrays matrix.
Detailed description of the invention
Fig. 1 is a kind of flow chart of the single frames coding structural light three-dimensional method for reconstructing based on multiple constraint of the present invention;
Fig. 2 is camera and projector in a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint of the present invention
Positional relationship coordinate diagram;
Fig. 3 is camera and projector in a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint of the present invention
Mounting structure schematic diagram;
Fig. 4 is DeBrujin sequence in a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint of the present invention
Schematic diagram;
Fig. 5 is the knot of element figure in a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint of the present invention
Structure schematic diagram.
Specific embodiment
The principle and features of the present invention will be described below with reference to the accompanying drawings, and the given examples are served only to explain the present invention, and
It is non-to be used to limit the scope of the invention.
As shown in Figure 1, a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint, includes the following steps,
S1 projects single frames coded structured light to testee surface by projector, and by the list of the projector projects
Frame coding structure light is tied on horizontal epipolar line;
S2 shoots the coding shadow of the single frames coded structured light deformed by the testee surface modulation using camera
Picture;
S3, using column direction constraint, horizontal epipolar line constraint and three kinds of the constraint relationships of coded sequence to the coded image into
Row decoding, obtains position of each encoded point in coded image in coded image;
S4 according to the system parameter demarcated in advance, and combines position of each encoded point in coded image in coded image,
Calculate the three dimensional point cloud on the testee surface.
A kind of key technology of the single frames coding structural light three-dimensional method for reconstructing based on multiple constraint of the present invention includes following
Several aspects:
1. the design based on horizontal epipolar line the constraint relationship
It is known as passing through the core face of this object space point, core face and image plane by plane made by photographic base and any object space point
Intersection be core line.According to photogrammetric and computer vision theory it is found that same place one is positioned at the same of stereogram
On name core line, this constraint relationship is known as horizontal epipolar line constraint.All core lines on inclination image are not parallel to each other, but are met at
Epipole.
On ideal image plane, all core lines are parallel to each other, and not only the core line in same image plane is parallel, but also image
Also parallel to upper corresponding core line, vertical parallax zero, this characteristic is particularly useful for stereopsis.The present invention in order to
The preferably coded image of the identification collected single frames structure light of camera, and reduce the calculating of the follow-up works such as mapping, matching
Two-dimensional search is reduced to one-dimensional linear search, improves arithmetic speed by amount, obtains ideal by carrying out reasonable design to hardware
Picture plane, it is intended to the structure light of projector projects is tied on horizontal epipolar line.
In Fig. 2, P is the projection plane (also referred to as image planes) of projector, and width W is highly H, and unit is pixel.To throw
Shadow planar central O is coordinate system XYZ, and O ' is projection centre, and OO ' is focal length f, and X-axis is the base direction of projector.In order to meet
The requirement of horizontal epipolar line will ensure that the X-axis of baseline and projector between camera and projector is overlapped on hardware installation, i.e. camera
Installation center C be located in the X-axis of projector, the projection plane center of projector and the installation center line OC (baseline) of camera
It is overlapped with the X-axis of shadow instrument;So the collected coded image of camera by core line correction after core line it is parallel, and in the row direction with
The coded image of projector projects corresponds.
In order to meet the requirement of horizontal epipolar line, the requirement on hardware installation is as follows:
(projector and camera, projector can be exposure mask projector to the instrument as used in the present invention, and camera can be
CCD camera) it is accurate optical measurement quasi-instrument, therefore optics sensor devices (including 2 imagings biographies in camera and projector
Sensor and 1 Inertial Measurement Unit) between absolute positional accuracy and stability camera later period measurement accuracy is played it is decisive
Factor, in order to guarantee that stablizing for camera and projector is indeformable, the sensor mounting framework structural member in camera and projector is adopted
With pack alloy roughing and finishing cooperation artificial aging (artificial aging specifically: first by structural member from mould furnace twice
In take out after, at room temperature stand 75 hours, discharge stress deformation;Then precision machine tool processing is carried out, and in open-air atmosphere
It is lower to stand 75 hours) guarantee that sensor mounting framework structural member deflection under the conditions of temperature change is small, and using high-precision
Five-axis machine tool processes each sensor installation position, guarantees that sensor is pacified using shaft hole matching and the multiple forms such as positioning that keep to the side
The consistency and stability of dress;Specifically as shown in Fig. 3 requirement, wherein projector 1 includes projector lens 11 and exposure mask piece 12;
Camera includes texture camera 2, point cloud camera 3 and auxiliary camera 4;Texture camera 2 includes texture camera camera lens 21 and optical filter 22;
Point cloud camera 3 includes point cloud camera lens 31 and optical center 32, and auxiliary camera 4 includes auxiliary camera camera lens 41;Sensor installs bone
Frame structural member 5 is equipped with IMU 51, and ((Inertial measurement unit, abbreviation IMU) is measurement object triaxial attitude angle
The device of (or angular speed) and acceleration, IMU are located at position of centre of gravity, guarantee that motion profile is accurate and error is small), sensor peace
Calibration has axis 52 on dress framing structural member 5.Also, sensor mounting framework structural member 5 is subtracted by four pieces during the installation process
Shake screw be fixed, suspension on the shell with guarantee acutely under conditions of vibration and impact under the conditions of, sensor module without
Negative effect guarantees that precision is effective for a long time.
2. improved DeBrujin sequence coding method
The two-dimensional expansion M-arrays matrix of coding method of the invention using DeBrujin sequence, two dimensional expansion
Exhibition M-arrays matrix is made of the stacking of multiple M-arrays matrixes, and each M-arrays matrix is by DeBrujin sequence
It constructs;The child window matrix H of each of M-arrays matrix m*n is in the entire two-dimensional expansion M-arrays
There is uniqueness in matrix, and
N=kmn- 1=R*C (1)
Wherein, N is the length of the DeBrujin sequence, and k is the number of element figure in the DeBrujin sequence, R
For the lateral resolution of the two-dimensional expansion M-arrays matrix, C is the longitudinal resolution of the two-dimensional expansion M-arrays matrix
Rate.
In treatment process the absolute of child window matrix H can be obtained by extracting the alignment characteristics of element in child window matrix H
Position encoded, this method only needs to project a frame grating, meet dynamic measure the needs of, if but element figure number it is few (k is small),
Measurement Resolution is low, and element figure number is more (k is big), and identification difficulty is big.
In order to solve above-mentioned contradiction, the present invention is improved, and measurement can be also reached in the case where pel number is few and is differentiated
The high purpose of rate.Only with two kinds of element figures in the present invention, i.e. k=2, with Binary Zero, 1 is indicated, reduces identification difficulty.
Simultaneously in order to expand the M-arrays matrix size, guarantee that Measurement Resolution, the present invention are fixed according to formula (1) in N value
Under the conditions of, R value is reduced, C value is increased, the M-arrays matrix that a R is 4 is constructed, improves longitudinal resolution (in list
In frame coding structure light, the higher the better for resolution ratio, i.e. R and C are the bigger the better, and in the present invention, we first reduce lateral resolution (R)
Guarantee longitudinal resolution (C), obtained the M-arrays matrix of only 4 rows, but laterally only has 4 rows to be much not achieved and answer
Repeat to encode in the row direction so using a strategy in invention with demand, every 4 row repeats a M-arrays
Matrix constitutes the two-dimensional expansion M-arrays matrix, to achieve the purpose that extend lateral resolution);Simultaneously based on level
The premise of core line constraint, laterally every 4 row repeats the M-arrays matrix, ensure that lateral resolution, as shown in figure 4, wherein
The content for including in entire big dotted line frame represents the entire two-dimensional expansion M-arrays matrix, each of in entire big dotted line frame
The content for including in small solid box represents a M-arrays matrix, and the content that the small dotted line frame in small solid box includes represents one
The child window matrix H of a m*n;Each M-arrays matrix is formed by DeBrujin sequence structure, entire two-dimensional expansion M-
Sequence in arrays matrix is the coded sequence of single frames coding structure.In decoding process, with present encoding point A (parent map
Member is 0 and 1 two binary character, corresponds to two attribute of width in coded image;Each of matrix element is corresponding to be compiled
Each characteristic point, i.e. encoded point in code image can be assigned a value of 0 or 1) by judging the width in region locating for encoded point for it
Centered on point, the child window matrix H of m*n is obtained according to upper and lower, left and right connection relationship, the child window matrix H is in entire institute
State in two-dimensional expansion M-arrays matrix it is unique, by m*n element forms in the child window matrix H binary sequence and its
Sequence in the M-arrays of place is matched, it may be determined that the encoded point A in the child window matrix H is in the M-
Position in arrays matrix, at once train value, line number of each encoded point in the M-arrays matrix in the child window matrix H
It is relative position for the entire two-dimensional expansion M-arrays matrix, each encoded point exists in the child window matrix H
Row number in the M-arrays matrix is absolute position, then benefit for the entire two-dimensional expansion M-arrays matrix
Each encoded point is removed to index in each child window matrix H in the two-dimensional expansion M-arrays matrix with horizontal epipolar line constraint
Absolute line number, and then obtain absolute coding position of each encoded point in the overall situation in each child window matrix H.
3. the design of single frames coded structured light
Present invention employs wide, narrow two kinds of element figures, to indicate binary one and 0, as shown in figure 5, according to both
Wide, narrow element figure combines the grid that above-mentioned coding method has obtained a kind of vertical narrow strips and horizontal circle angular width striped intersects composition
Grid pattern.Vertical narrow strips link together, and naturally occur straight line (center lines of i.e. vertical narrow strips), this straight line with
Horizontal epipolar line intersects to obtain characteristic point, i.e. encoded point;Encoded point row number on the center line of vertical narrow strips is identical, therefore, this
By extracting the center line of vertical narrow strips in invention, the constraint relationship of encoded point in a column direction can be obtained, and (i.e. column direction is about
Beam relationship, there are two levels for column direction the constraint relationship: first is that the upper and lower connection relationship of encoded point and its neighborhood point, with after an action of the bowels
The child window matrix H of vertical m*n is continued, second is that the encoded point row number on the center line of the vertical narrow strips of same is identical, with after an action of the bowels
Continuous point cloud encryption).In decoding process, it can determine the upper and lower connection relationship of encoded point Yu its neighborhood point by center line, pass through water
Flat core line constraint can determine the left and right connection relationship of encoded point Yu its neighborhood point, fast and accurately establish the child window matrix of m*n
H obtains position of the encoded point in part (M-arrays matrix), and because same place is tied to same a line by horizontal epipolar line
On, position of the encoded point in global (two-dimensional expansion M-arrays matrix) can be obtained, improve decoded stability and robust
Property.
Point cloud encryption (sub-dense points, sub- dense point cloud extract) can be parsed out vertical by being expressed above
Code value corresponding to each encoded point on the center line of narrow strips, is then compared with the coded image of projector projects
It is right, the corresponding parallax value of each encoded point can be calculated.And in fact, based on minimum planes it is assumed that in vertical narrow strips
(including non-coding point) corresponding parallax of all pixels point should be smooth distribution, i.e., neighbour on the center line of vertical narrow strips on heart line
Nearly pixel parallax is close.Pixel on the center line of all vertical narrow strips is handled, Dynamic Programming, confidence are utilized
Derive optimization method can be parsed out in all vertical narrow strips using the difference minimum of neighborhood pixels parallax as objective function
Corresponding parallax value in corresponding pixel points on heart line.Compared to other methods, effectively raise single frames structure light parsing
Scanning resolution and efficiency.
In addition, the precision of encoded point position positioning can also largely effect on the precision of reconstruction, horizontal width striped is using circle
Angular shape, is presented oval form in the collected image of camera, it is subsequent can be combined by sub-pixel edge positioning method it is ellipse
The accurate position for extracting encoded point of circle fitting, improves measurement accuracy.
The beneficial effects of the present invention are:
(1) in order to preferably extend and identify structure light coding image, and two-dimensional search is reduced to one-dimensional linear search,
Present invention proposition constrains in structure light coding image on horizontal epipolar line, is conducive to line direction and indexes, enormously simplifies answering for problem
Miscellaneous degree simultaneously improves operation efficiency.
(2) present invention expands the ranks of two-dimensional expansion M-arrays matrix on the basis of only there are two pel
Exhibition reduces identification difficulty and improves Measurement Resolution.
(3) the invention proposes a kind of vertical narrow strips and horizontal circle angular width striped to intersect the Lattice encoding pattern constituted,
Same row symbol, by extracting the center line of narrow strips, can be obtained the constraint of symbol in a column direction and close in same narrow strips
System provides more information for subsequent processing, accelerates decoding efficiency, improves robustness.Meanwhile can also by center line to cloud into
Row encryption, improves Measurement Resolution.In addition, rounded shapes would be even more beneficial to the accurate extraction that sub-pixel is carried out to grid point,
Improve measurement accuracy.
In addition, the present disclosure applies equally to double camera structured light three-dimensional measurement systems.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and
Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (9)
1. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint, it is characterised in that: include the following steps,
S1 projects single frames coded structured light to testee surface by projector, and the single frames of the projector projects is compiled
Code structure light is tied on horizontal epipolar line;
S2 shoots the coded image of the single frames coded structured light deformed by the testee surface modulation using camera;
S3 solves the coded image using column direction constraint, horizontal epipolar line constraint and three kinds of the constraint relationships of coded sequence
Code, obtains position of each encoded point in the coded image in the coded image;
S4, according to the system parameter demarcated in advance, and in conjunction with each encoded point in the coded image in the coded image
Position calculates the three dimensional point cloud on the testee surface.
2. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 1, feature
Be: in the S1, the single frames coded structured light of the projector projects, which is tied to the specific method on horizontal epipolar line, is,
By putting the projector and the camera around the testee, make between the camera and the projector
Baseline be overlapped with the X-axis of the projector;
Wherein, the baseline between the camera and the projector is specifically, the installation center of the camera and the projector
Projection plane center between line;The X-axis of the projector is specifically, be original with the projection plane center of the projector
Point makees XYZ coordinate system, and the X-axis in XYZ coordinate system is the X-axis of the projector.
3. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 1 or 2, special
Sign is: the coding method of the single frames coded structured light of the projector projects using DeBrujin sequence two-dimensional expansion
M-arrays matrix;
The two-dimensional expansion M-arrays matrix is made of the stacking of multiple M-arrays matrixes, and each M-arrays matrix
It is formed by DeBrujin sequence structure;
The child window matrix H of each of M-arrays matrix m*n is in the entire two-dimensional expansion M-arrays matrix
In have uniqueness, and, N=kmn- 1=R*C, wherein N is the length of the DeBrujin sequence, and k is the DeBrujin sequence
The number of element figure in column, R are the lateral resolution of the two-dimensional expansion M-arrays matrix, and C is the two-dimensional expansion M-
The longitudinal resolution of arrays matrix.
4. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 3, feature
Be: the line number of the M-arrays matrix is 4, then the R=4.
5. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 3 or 4, special
Sign is: including wide, narrow two kinds of element figures in the single frames coded structured light of the projector projects, indicates two to respectively correspond
1 in system and 0;
According to wide, narrow two kinds of element figures, and the coding method of the two-dimensional expansion M-arrays matrix of DeBrujin sequence is combined,
It obtains a kind of vertical narrow strips and intersects the comb mesh pattern constituted with horizontal wide striped, the comb mesh pattern is the projector throwing
The pattern that the single frames coded structured light of injection is formed.
6. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 5, feature
Be: the horizontal wide striped is specially horizontal circle angular width striped.
7. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 5, feature
Be: the S3 specifically,
S31 extracts the encoded point in the coded image, by column direction the constraint relationship and horizontal epipolar line the constraint relationship described
The upper and lower, left and right connection relationship between the encoded point is established in two-dimensional expansion M-arrays matrix, obtains multiple sons
Window matrix H;
The M-arrays matrix at each child window matrix H and place is carried out encoding array matching respectively, really by S32
Position of each encoded point in the M-arrays matrix in fixed each child window matrix H;
S33 indexes out in each child window matrix H each encoded point in the two dimension using horizontal epipolar line the constraint relationship respectively
Extend the position in M-arrays matrix.
8. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 7, feature
It is: in the S31, by extracting the center line of vertical narrow strips, obtains the column direction the constraint relationship of encoded point.
9. a kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint according to claim 7 or 8, special
Sign is: in the S32, position of each encoded point in the M-arrays matrix is distinguished in the child window matrix H
Including line number and row number;Line number of each encoded point in the M-arrays matrix is relative to entire in the child window matrix H
It is relative position for the two-dimensional expansion M-arrays matrix, each encoded point is in the M- in the child window matrix H
Row number in arrays matrix is absolute position for the entire two-dimensional expansion M-arrays matrix;
In the S33, each encoded point in each child window matrix H is indexed out respectively using horizontal epipolar line the constraint relationship and is existed
Absolute line number in the two-dimensional expansion M-arrays matrix.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810980173.2A CN109242957A (en) | 2018-08-27 | 2018-08-27 | A kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810980173.2A CN109242957A (en) | 2018-08-27 | 2018-08-27 | A kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109242957A true CN109242957A (en) | 2019-01-18 |
Family
ID=65068291
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810980173.2A Withdrawn CN109242957A (en) | 2018-08-27 | 2018-08-27 | A kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109242957A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109993826A (en) * | 2019-03-26 | 2019-07-09 | 中国科学院深圳先进技术研究院 | A kind of structural light three-dimensional image rebuilding method, equipment and system |
CN110310365A (en) * | 2019-06-27 | 2019-10-08 | 四川大学 | A kind of three-dimensional rebuilding method and device |
CN111023996A (en) * | 2019-11-18 | 2020-04-17 | 西安电子科技大学 | Single-frame dynamic three-dimensional measurement method |
CN111325831A (en) * | 2020-03-04 | 2020-06-23 | 中国空气动力研究与发展中心超高速空气动力研究所 | Color structured light bar detection method based on hierarchical clustering and belief propagation |
CN111336950A (en) * | 2020-04-11 | 2020-06-26 | 武汉玄景科技有限公司 | Single-frame measuring method and system combining spatial coding and line structure light |
CN111336949A (en) * | 2020-04-11 | 2020-06-26 | 武汉玄景科技有限公司 | Spatial coding structured light three-dimensional scanning method and system |
CN112070842A (en) * | 2020-07-28 | 2020-12-11 | 安徽农业大学 | Multi-camera global calibration method based on orthogonal coding stripes |
CN114299172A (en) * | 2021-12-31 | 2022-04-08 | 广东工业大学 | Planar coding target for visual system and real-time pose measurement method thereof |
WO2023082816A1 (en) * | 2021-11-15 | 2023-05-19 | 资阳联耀医疗器械有限责任公司 | Structured light coding method and system for three-dimensional information reconstruction |
-
2018
- 2018-08-27 CN CN201810980173.2A patent/CN109242957A/en not_active Withdrawn
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109993826B (en) * | 2019-03-26 | 2023-06-13 | 中国科学院深圳先进技术研究院 | Structured light three-dimensional image reconstruction method, device and system |
CN109993826A (en) * | 2019-03-26 | 2019-07-09 | 中国科学院深圳先进技术研究院 | A kind of structural light three-dimensional image rebuilding method, equipment and system |
CN110310365A (en) * | 2019-06-27 | 2019-10-08 | 四川大学 | A kind of three-dimensional rebuilding method and device |
CN111023996B (en) * | 2019-11-18 | 2021-05-14 | 西安电子科技大学 | Single-frame dynamic three-dimensional measurement method |
CN111023996A (en) * | 2019-11-18 | 2020-04-17 | 西安电子科技大学 | Single-frame dynamic three-dimensional measurement method |
CN111325831B (en) * | 2020-03-04 | 2022-07-01 | 中国空气动力研究与发展中心超高速空气动力研究所 | Color structured light bar detection method based on hierarchical clustering and belief propagation |
CN111325831A (en) * | 2020-03-04 | 2020-06-23 | 中国空气动力研究与发展中心超高速空气动力研究所 | Color structured light bar detection method based on hierarchical clustering and belief propagation |
CN111336949A (en) * | 2020-04-11 | 2020-06-26 | 武汉玄景科技有限公司 | Spatial coding structured light three-dimensional scanning method and system |
CN111336950A (en) * | 2020-04-11 | 2020-06-26 | 武汉玄景科技有限公司 | Single-frame measuring method and system combining spatial coding and line structure light |
CN112070842A (en) * | 2020-07-28 | 2020-12-11 | 安徽农业大学 | Multi-camera global calibration method based on orthogonal coding stripes |
CN112070842B (en) * | 2020-07-28 | 2023-03-21 | 安徽农业大学 | Multi-camera global calibration method based on orthogonal coding stripes |
WO2023082816A1 (en) * | 2021-11-15 | 2023-05-19 | 资阳联耀医疗器械有限责任公司 | Structured light coding method and system for three-dimensional information reconstruction |
CN114299172A (en) * | 2021-12-31 | 2022-04-08 | 广东工业大学 | Planar coding target for visual system and real-time pose measurement method thereof |
CN114299172B (en) * | 2021-12-31 | 2022-07-08 | 广东工业大学 | Planar coding target for visual system and real-time pose measurement method thereof |
US11699244B2 (en) | 2021-12-31 | 2023-07-11 | Guangdong University Of Technology | Planar coding target for vision system and real-time pose measurement method thereof |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109242957A (en) | A kind of single frames coding structural light three-dimensional method for reconstructing based on multiple constraint | |
CN107063129B (en) | A kind of array parallel laser projection three-dimensional scan method | |
CN104335005B (en) | 3D is scanned and alignment system | |
US10152800B2 (en) | Stereoscopic vision three dimensional measurement method and system for calculating laser speckle as texture | |
CN104541127B (en) | Image processing system and image processing method | |
CN110487216A (en) | A kind of fringe projection 3-D scanning method based on convolutional neural networks | |
US8836766B1 (en) | Method and system for alignment of a pattern on a spatial coded slide image | |
CN110288642A (en) | Three-dimension object fast reconstructing method based on camera array | |
CN112525107B (en) | Structured light three-dimensional measurement method based on event camera | |
CN107967697B (en) | Three-dimensional measurement method and system based on color random binary coding structure illumination | |
CN111028295A (en) | 3D imaging method based on coded structured light and dual purposes | |
CN104596439A (en) | Speckle matching and three-dimensional measuring method based on phase information aiding | |
CN111563952B (en) | Method and system for realizing stereo matching based on phase information and spatial texture characteristics | |
CN112945141B (en) | Structured light rapid imaging method and system based on micro-lens array | |
CN107990846B (en) | Active and passive combination depth information acquisition method based on single-frame structured light | |
CN108613637A (en) | A kind of structured-light system solution phase method and system based on reference picture | |
Furukawa et al. | One-shot entire shape acquisition method using multiple projectors and cameras | |
JP5761750B2 (en) | Image processing method and apparatus | |
CN110349257A (en) | A kind of binocular measurement missing point cloud interpolating method based on the mapping of phase puppet | |
CN104634323B (en) | A kind of multistage shooting tri-dimensional photographing system and method | |
CN110248179B (en) | Camera pupil aberration correction method based on light field coding | |
CN108895979A (en) | The structure optical depth acquisition methods of line drawing coding | |
CN107063131B (en) | A kind of time series correlation non-valid measurement point minimizing technology and system | |
CN115290004A (en) | Underwater parallel single-pixel imaging method based on compressed sensing and HSI | |
CN111023999B (en) | Dense point cloud generation method based on spatial coding structured light |
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 | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190118 |
|
WW01 | Invention patent application withdrawn after publication |