CN109285213A - Comprehensive polarization three-dimensional rebuilding method - Google Patents
Comprehensive polarization three-dimensional rebuilding method Download PDFInfo
- Publication number
- CN109285213A CN109285213A CN201810787882.9A CN201810787882A CN109285213A CN 109285213 A CN109285213 A CN 109285213A CN 201810787882 A CN201810787882 A CN 201810787882A CN 109285213 A CN109285213 A CN 109285213A
- Authority
- CN
- China
- Prior art keywords
- polarization
- axis
- polarization image
- image
- scene
- 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
- 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
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/50—Lighting effects
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Graphics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The present invention relates to a kind of comprehensive polarization three-dimensional rebuilding methods, comprising: obtains the target polarization image and luminous intensity for rebuilding the first side of object;Obtain the target polarization image and luminous intensity for rebuilding object second side opposite with the first side;First surface normal is calculated according to the target polarization of the first side and luminous intensity, and first surface normal is corrected;Second surface normal is calculated according to the target polarization of second side and luminous intensity, and second surface normal is corrected;The first side 3D shape and second side 3D shape are obtained respectively using the first surface normal after correction and the second surface normal after correction;First side 3D shape and second side 3D shape are merged, the target three-dimensional image for rebuilding object is obtained.By the present invention in that with camera and mirror system, carry out three-dimensional reconstruction using polarization information, realize faster, the higher three-dimensional imaging of precision.
Description
Technical field
The invention belongs to technical field of image processing, and in particular to a kind of comprehensive polarization three-dimensional rebuilding method.
Background technique
Three-dimensional reconstruction refers to the three-dimensional mathematical model established to the three dimensional object to be imaged and be suitble to computer representation and processing.
By three-dimensional reconstruction, the data such as surface profile, details, the depth of object can be by Precise Representation.With this technology
Demand is more and more extensive, and the application prospect in fields such as historical relic's protection, medicine, virtual display and ergonomic designs,
The demand that body surface three-dimensional is imaged in people has been not only to carry out three-dimensional to the surface data in object a direction
It rebuilds, needs to carry out object comprehensive three-dimensional reconstruction in more application scenarios.
Currently, being able to carry out the common methods of omnidirectional three-dimensional reconstruction has the three-dimensional reconstruction side based on 3 D laser scanning
Method, three-dimensional rebuilding method based on structure light scan etc..Three-dimensional rebuilding method based on 3 D laser scanning is linear by receiving
The linear laser that laser projects is irradiated to reflected optical signal on barrier, realizes the three-dimensional reconstruction of body surface.
The imaging system of this method has certain improvement to the laser imaging apparatus of previous valuableness, and the detail data of reconstructed results is more rich
Richness has certain application in three-dimensional reconstruction field.
But existing three-dimensional rebuilding method needs to rotate object or require imaging system to carry out around object
Rotation causes system structure complicated, and operation time is long, and operating distance is short or reconstructed results detail data loses the originals such as serious
Cause strongly limits the application range of three-dimensional reconstruction.In addition, needing during carrying out data acquisition to target by hard
Part equipment and circuit control, to target per one group of data is once acquired, so the time is long in acquisition data procedures, in addition
Since the error of hardware device is it is also possible to cause the decline of precision.
Summary of the invention
In order to solve the above-mentioned problems in the prior art, the present invention provides a kind of comprehensive polarization three-dimensional reconstruction sides
Method.The technical problem to be solved in the present invention is achieved through the following technical solutions:
The present invention provides a kind of comprehensive polarization three-dimensional rebuilding methods, comprising:
S1: the luminous intensity of the target polarization image and the first side of rebuilding the first side of object is obtained;
S2: the luminous intensity of the target polarization image and second side of rebuilding object second side opposite with the first side is obtained;
S3: first surface normal is calculated according to the luminous intensity of the target polarization of the first side and the first side, and to described first
Surface normal is corrected;
S4: second surface normal is calculated according to the luminous intensity of the target polarization of second side and second side, and to described second
Surface normal is corrected;
S5: the first side three-dimensional shaped is obtained respectively using the first surface normal after correction and the second surface normal after correction
Shape and second side 3D shape;
S6: merging first side 3D shape and second side 3D shape, obtains the mesh for rebuilding object
Mark 3-D image.
In one embodiment of the invention, the S1 includes:
S11: under light source irradiation, the image data of the first side of object is rebuild in acquisition, obtains the first scene polarization image
I1';
S12: three times by the rotation of polarizing film sequence, 45 ° of rotation every time acquires the picture number of first side every time after rotation
According to obtaining the second scene polarization image I respectively2', third scene polarize I3' and the 4th scene polarization image I4';
S13: to the first scene polarization image I1', the second scene polarization image I2', third scene polarize I3' and the
Four scene polarization image I4' be split respectively, obtain first object polarization image I1, the second target polarization image I2, third mesh
Mark polarization image I3With the 4th target polarization image I4;
S14: pass through the second scene polarization image I2' and the 4th scene polarization image I4' obtain the luminous intensity of first side
I0。
In one embodiment of the invention, the S2 includes:
S21: by two reflecting mirrors, the image data of second side of object backlight is rebuild in acquisition, and to two reflecting mirrors
Image data carry out Feature Points Matching, obtain the 5th scene polarization image I5';
S22: three times by the rotation of polarizing film sequence, 45 ° of rotation every time acquires the picture number of described second side every time after rotation
According to, respectively obtain the 6th scene polarization image I6', the 7th scene polarize I7' and the 8th scene polarization image I8';
S23: to the 5th scene polarization image I5', the 6th scene polarization image I6', the 7th scene polarize I7' and the
Eight scene polarization image I8' be split respectively, obtain the 5th target polarization image I5, the 6th target polarization image I6, the 7th mesh
Mark polarization image I7With the 8th target polarization image I8;
S24: pass through the 6th scene polarization image I6' and the 8th scene polarization image I8' obtain the luminous intensity of described second side
I0’。
In one embodiment of the invention, the S3 includes:
S31: the degree of polarization P of the first side surface is calculated;
S32: the azimuth of the first side surface is calculatedWith the incidence angle θ of the first side surface;
S33: according to the azimuthThe surface normal of first side is calculated with the incidence angle θIn x-axis
Component nxWith the component n in y-axisy;
S34: the surface normal is calculatedCorrection coefficient in x-axisWith the correction coefficient in y-axis
S35: pass through correction coefficientWithThe correction component n in x-axis is obtained respectivelykxAnd the correction component in y-axis
nky。
In one embodiment of the invention, the S31 includes:
Utilize formulaCalculate the degree of polarization P of first side surface, wherein I, Q and U indicate to rebuild object
The Stokes vector of the first side of body, its calculation formula is:
Wherein, ExIndicate the vector of electric field in the direction of the x axis, EyIndicate the vector of electric field in the y-axis direction;ILIt is inclined for circle
The left polarization image of vibration;IRFor the right polarization image of circular polarization.
In one embodiment of the invention, the S32 includes:
S321: first object polarization image I is utilized1, the second target polarization image I2, third target polarization image I3With
Four target polarization image I4Calculate the azimuth of first side surface
S322: the incidence angle θ of first side surface is calculated according to the degree of polarization P.
In one embodiment of the invention, the S33 includes:
According to the azimuthThe surface normal of first side is calculated with the incidence angle θComponent n in x-axisx
With the component n in y-axisy, calculation formula are as follows:
In one embodiment of the invention, the S34 includes:
S341: being based on Lambertian reflection model, and the simulation normal for rebuilding body surface is established using shape from shading method
Analogue component n in x-axis and y-axissxAnd nsy, calculation formula are as follows:
Wherein, η is body surface albedo, and L is light source direction, LxAnd LyRespectively light source direction is in x-axis and y-axis direction
On component, I0For the luminous intensity of the first side;
S342: the surface normal is calculatedCorrection coefficient in x-axisCalculation formula are as follows:
S343: the surface normal is calculatedCorrection coefficient on the y axisCalculation formula are as follows:
In one embodiment of the invention, the S35 includes:
S351: according to correction coefficientAnd surface normalComponent n in x-axisxX-axis component n after calculating correctioncx,
Calculation formula are as follows:
S352: according to correction coefficientAnd surface normalComponent n on the y axisyY-axis component n after calculating correctioncy,
Calculation formula are as follows:
In one embodiment of the invention, the S5 includes:
S51: the x-axis component n after the correction of first surface normal is utilizedcxWith the y-axis component n after correctioncyCalculate the first side three
Tie up shape, calculation formula are as follows:
Cos (z)=∫ ∫ ((zx-ncx)2+(zy-ncy)2) dxdy,
Wherein, z indicates to rebuild the three-dimensional surface of object, zxIndicate the gradient of the three-dimensional surface of object in the direction of the x axis, zy
Indicate the gradient of the three-dimensional surface of object in the y-axis direction;
S52: second side three-dimensional shaped is calculated using the x-axis component after the correction of second surface normal and the y-axis component after correction
Shape.
Compared with prior art, the beneficial effects of the present invention are:
1, the comprehensive polarization three-dimensional rebuilding method of the present invention utilizes polarization by using one camera and mirror system
Information carries out three-dimensional reconstruction, is effectively simplified in conventional three-dimensional method for reconstructing that the device is complicated, target information obtains low efficiency etc. and asks
Topic, realize faster, the higher three-dimensional imaging of precision.
2, the present invention can be simply square using comprehensive polarization three-dimensional rebuilding method of the one camera in conjunction with mirror system
Just target three-dimensional reconstruction is carried out, the complexity of system structure is reduced, while reducing the mistake introduced in multiple information measurement
Difference.
Detailed description of the invention
Fig. 1 is the present invention for obtaining the structural schematic diagram for rebuilding the dark box device of object polarization image and depth information;
Fig. 2 is a kind of flow chart of comprehensive polarization three-dimensional rebuilding method provided in an embodiment of the present invention;
Fig. 3 is the relationship signal that the present invention rebuilds object incident light beam strikes angle, incident orientation angle and body surface normal
Figure;
Fig. 4 is that the light intensity being fitted using least square method rotates the cosine curve figure of angle change with polarizing film.
Specific embodiment
The content of present invention is described further combined with specific embodiments below, but embodiments of the present invention are not limited to
This.
Embodiment one
Referring to Figure 1, Fig. 1 is that the structure of device of the present invention for obtaining reconstruction object polarization image and depth information is shown
It is intended to.First integral ball 1, detector 2, polarizing film 3 including sequence setting in the dark box device, reconstruction object 4, plane are anti-
Penetrate mirror assembly 5 and second integral ball 6, wherein first integral ball 1, second integral ball 6, detector 2 and polarizing film 3 are fixed
On optical manipulation platform;Position of the object 4 between plane reflection mirror assembly 5 and detector 2 is rebuild in adjustment, makes plane reflection
The information for rebuilding 4 back side of object can be got in mirror assembly 5 simultaneously;Polarizing film 3 be fixed on detector 2 and rebuild object 4 it
Between, and polarizing film 3 is parallel with the camera lens of detector 2;First integral ball 1 and second integral ball 6 are respectively the two sides for rebuilding object 4
Light source illumination is provided.In the present embodiment, plane reflection mirror assembly 5 includes two pieces of plane mirrors at an angle, wherein
Angle between two pieces of plane mirrors is adjusted according to different article size and operating distance etc., guarantees two planes
The overlapping image of some in reflecting mirror, to obtain rebuilding the complete image of object.
Detector 2 in the present embodiment uses CMOS camera, and camera bellows is square, and camera bellows inner wall covers black-out cloth, is used to
It absorbs first integral ball 1 and second integral ball 6 is radiated at the light on carton inner wall, polarizing film 3 uses linear polarizer.
Referring to figure 2., Fig. 2 is a kind of flow chart of comprehensive polarization three-dimensional rebuilding method provided in an embodiment of the present invention.
The three-dimensional rebuilding method of the present embodiment includes the following steps:
S1: the luminous intensity of the target polarization image and the first side of rebuilding the first side of object is obtained;
S2: the light intensity of the target polarization image for rebuilding object second side opposite with the first side and second side is obtained
Degree;
S3: first surface normal is calculated according to the luminous intensity of the target polarization of the first side and the first side, and to described first
Surface normal is corrected;
S4: second surface normal is calculated according to the luminous intensity of the target polarization of second side and second side, and to described second
Surface normal is corrected;
S5: the first side three-dimensional shaped is obtained respectively using the first surface normal after correction and the second surface normal after correction
Shape and second side 3D shape;
S6: merging first side 3D shape and second side 3D shape, obtains the reconstruction object
Target three-dimensional image.
Further, the S1 includes:
S11: under light source irradiation, the image data of the first side of object is rebuild in acquisition, obtains the first scene polarization image
I1';
Specifically, under the irradiation of first integral ball 1, using imaging detector CMOS camera acquisition body surface close to the
The reflected light of one integrating sphere, 1 side, to obtain the first scene polarization image I1'.Here, the first side indicates to rebuild body surface
Close to the side of first integral ball 1.
S12: three times by the rotation of polarizing film sequence, 45 ° of rotation every time acquires the picture number of first side every time after rotation
According to obtaining the second scene polarization image I respectively2', third scene polarization image I3' and the 4th scene polarization image I4';
Specifically, polarizing film is rotated 45 °, using imaging detector CMOS camera acquisition body surface close to first integral
The reflected light of 1 side of ball, to obtain the second scene polarization image I2';Polarizing film is rotated to 45 ° (that is, 90 ° of corotatings) again,
Body surface is acquired close to the reflected light of 1 side of first integral ball, to obtain third scene polarization image I3';By polarizing film
It rotates again 45 ° (that is, 135 ° of corotatings), acquires body surface close to the reflected light of 1 side of first integral ball, to obtain the 4th
Scene polarization image I4’。
S13: to the first scene polarization image I1', the second scene polarization image I2', third scene polarize I3' and the
Four scene polarization image I4' be split respectively, obtain first object polarization image I1, the second target polarization image I2, third mesh
Mark polarization image I3With the 4th target polarization image I4;
To reduce calculation amount, need to be split the object scene polarization image got respectively, by object and background
It separates, removes background parts.In the present embodiment, using Threshold Segmentation Algorithm by the object and background segment in polarization image
It opens, first object polarization image I is obtained after Threshold segmentation1, the second target polarization image I2, third target polarization image I3With
Four target polarization image I4。
S14: pass through the second scene polarization image I2' and the 4th scene polarization image I4' obtain the luminous intensity of first side
I0。
Specifically, according to polarization characteristic, the orthogonal information sum of two polarization azimuths is equal to light intensity value, in this reality
It applies in example, passes through I0=I2+I4Calculate the intensity information for rebuilding the first side surface of object.
Further, the S2 includes:
S21: by two reflecting mirrors, the image data of second side of object backlight is rebuild in acquisition, and to two reflecting mirrors
Image data carry out Feature Points Matching, obtain the 5th scene polarization image I5';
It, can be anti-by obtaining reflecting mirror since the light polarization after reflecting mirror reflects will not change
The optical information penetrated generate rebuild object backlight second side image data.Specifically, detector is adjusted, keeps it flat to two
Reconstruction object in the reflecting mirror of face is imaged simultaneously, since the image in two reflecting mirrors has lap, to reflecting mirror
In image by SIFT (Scale invariant features transform) matching process carry out Feature Points Matching, to obtain the 5th of second side
Scene polarization image I5'.Herein, second side indicates to rebuild body surface close to the side of second integral ball 6.
S22: three times by the rotation of polarizing film sequence, 45 ° of rotation every time acquires the picture number of described second side every time after rotation
According to, respectively obtain the 6th scene polarization image I6', the 7th scene polarize I7' and the 8th scene polarization image I8';
Specifically, polarizing film is rotated 45 °, is reflected using the imaging detector CMOS camera acquisition object back side through reflecting mirror
Information afterwards carries out Feature Points Matching by SIFT matching process to the image in reflecting mirror, to obtain the 6th of second side
Scene polarization image I6', similarly, then polarizing film is rotated 45 ° respectively, obtains the 7th scene polarization I7 ' and the 8th scene is inclined
Shake image I8 '.
S23: to the 5th scene polarization image I5', the 6th scene polarization image I6', the 7th scene polarize I7' and the
Eight scene polarization image I8' be split respectively, obtain the 5th target polarization image I5, the 6th target polarization image I6, the 7th mesh
Mark polarization image I7With the 8th target polarization image I8;
In the present embodiment, using Threshold Segmentation Algorithm by polarization image object and background segment open.
S24: pass through the 6th scene polarization image I6' and the 8th scene polarization image I8' obtain the luminous intensity of described second side
I0’。
In the present embodiment, pass through I0'=I6’+I8' calculate the intensity information for rebuilding the second side surface of object.
The comprehensive polarization three-dimensional rebuilding method of the present embodiment utilizes polarization by using one camera and mirror system
Information carries out three-dimensional reconstruction, is effectively simplified in conventional three-dimensional method for reconstructing that the device is complicated, target information obtains low efficiency etc. and asks
Topic, realize faster, the higher three-dimensional imaging of precision.
Embodiment two
On the basis of the above embodiments, the present embodiment is to the three-dimensional reconstruction step in comprehensive polarization three-dimensional rebuilding method
Calculating process be described in detail.
Specifically, the S3 includes:
S31: the degree of polarization P of the first side surface is calculated;
Specifically, formula is utilizedCalculate the degree of polarization P of first side surface, wherein I, Q and U table
Show the Stokes vector for rebuilding the first side of object, its calculation formula is:
Wherein, ExIndicate the vector of electric field in the direction of the x axis, EyIndicate the vector of electric field in the y-axis direction;ILIt is inclined for circle
The left polarization image of vibration;IRFor the right polarization image of circular polarization.
In addition, calculating degree of polarization P, there are also a kind of alternative solutions.Fig. 4 is referred to, Fig. 4 is fitted using least square method
Light intensity rotates the cosine curve figure of angle change with polarizing film.Utilize first object polarization image I1, the second target polarization image
I2, third target polarization image I3With the 4th target polarization image I4, light intensity I, which is fitted, using least square method revolves with polarizing film
Gyration θpolThe cosine curve of variation, as shown in Figure 4.The maxima and minima I of available light intensity I from Fig. 4maxWith
Imin, it can be obtained by degree of polarization P further according to the definition of degree of polarization, its calculation formula is:
S32: the azimuth of the first side surface is calculatedAnd incidence angle θ;
Fig. 3 is referred to, Fig. 3 is that the present invention rebuilds object incident light beam strikes angle, incident orientation angle and body surface normal
Relation schematic diagram.Specifically, first object polarization image I is utilized1, the second target polarization image I2, third target polarization image
I3With the 4th target polarization image I4, azimuth is calculated according to Stokes vector (a kind of representation method of polarization state)
Calculation formula are as follows:
Then, incidence angle θ, calculation formula can be calculated according to the relationship of degree of polarization P and incidence angle θ are as follows:
Wherein, n indicates the refractive index of body surface, and in the present embodiment, generally taking body surface refractive index is 1.5.
S33: according to the azimuthThe surface normal of first side is calculated with the incidence angle θPoint in x-axis
Measure nxWith the component n in y-axisy, calculation formula are as follows:
S34: the surface normal is calculatedCorrection coefficient in x-axisWith the correction coefficient in y-axis
Specifically, the S34 includes:
S341: being based on Lambertian reflection model, and the simulation normal for rebuilding body surface is established using shape from shading method
Analogue component n in x-axis and y-axissxAnd nsy, calculation formula are as follows:
Wherein, η is body surface albedo, and L is light source direction, LxAnd LyRespectively light source direction is in x-axis and y-axis direction
On component, I0For the luminous intensity of the first side;
S342: the surface normal is calculatedCorrection coefficient in x-axisCalculation formula are as follows:
S343: the surface normal is calculatedCorrection coefficient on the y axisCalculation formula are as follows:
S35: pass through correction coefficientWithThe correction component n in x-axis is obtained respectivelykxAnd the correction component in y-axis
nky。
Specifically, the S35 includes:
S351: according to correction coefficientAnd surface normalComponent n in x-axisxX-axis component n after calculating correctioncx,
Calculation formula are as follows:
S352: according to correction coefficientAnd surface normalComponent n on the y axisyY-axis component n after calculating correctioncy,
Calculation formula are as follows:
It is similar with step S3, step S4 the following steps are included:
S41: the degree of polarization P' of the second side surface is calculated;
Specifically, specifically, formula is utilizedCalculate the degree of polarization P' of second side surface, wherein
I, Q and U indicates to rebuild the Stokes vector of object second side, its calculation formula is:
Wherein, ExIndicate the vector of electric field in the direction of the x axis, EyIndicate the vector of electric field in the y-axis direction;ILIt is inclined for circle
The left polarization image of vibration;IRFor the right polarization image of circular polarization.
S32: the azimuth of the second side surface is calculatedWith incidence angle θ ';
Specifically, first object polarization image I is utilized5, the second target polarization image I6, third target polarization image I7With
4th target polarization image I8, azimuth is calculated according to Stokes vector (a kind of representation method of polarization state)Meter
Calculate formula are as follows:
Then, according to degree of polarization P' and incidence angle θ ' relationship can be calculated incidence angle θ ', calculation formula are as follows:
Wherein, n indicates the refractive index of body surface, and in the present embodiment, generally taking body surface refractive index is 1.5.
S43: according to the azimuthWith the surface normal of the incidence angle θ ' calculating first sideIn x-axis
Component nx' and y-axis on component ny', calculation formula are as follows:
S44: the surface normal is calculatedCorrection coefficient in x-axisWith the correction coefficient in y-axis
Specifically, S44 includes:
S441: being based on Lambertian reflection model, and the simulation normal for rebuilding body surface is established using shape from shading method
Analogue component n in x-axis and y-axissx' and nsy', calculation formula are as follows:
Wherein, η is body surface albedo, and L is light source direction, LxAnd LyRespectively light source direction is in x-axis and y-axis direction
On component, I0' be second side luminous intensity;
S442: the surface normal is calculatedCorrection coefficient in x-axisCalculation formula are as follows:
S443: the surface normal is calculatedCorrection coefficient on the y axisCalculation formula are as follows:
S45: pass through correction coefficientWithThe correction component n in x-axis is obtained respectivelykx' and y-axis on correction point
Measure nky'。
Specifically, the S45 includes:
S451: according to correction coefficientAnd surface normalComponent n in x-axisx' calculate the x-axis component after correction
ncx', calculation formula are as follows:
S452: according to correction coefficientAnd surface normalComponent n on the y axisy' calculate the y-axis component after correction
ncy', calculation formula are as follows:
Further, the S5 includes:
S51: the x-axis component n after the correction of first surface normal is utilizedcxWith y-axis component ncyThe first side 3D shape is calculated,
Calculation formula are as follows:
Cos (z)=∫ ∫ ((zx-ncx)2+(zy-ncy)2) dxdy,
Wherein, z indicates to rebuild the first side three-dimensional surface of object, zxIndicate the first side three-dimensional surface of object in x-axis side
Upward gradient, zyIndicate the gradient of the first side three-dimensional surface of object in the y-axis direction;
S52: x-axis component and y-axis component after being corrected using second surface normal calculate second side 3D shape, calculate public
Formula are as follows:
Cos (z')=∫ ∫ ((zx'-ncx')2+(zy'-ncy')2) dxdy,
Wherein, z' indicates to rebuild second side three-dimensional surface of object, zx' indicate second side three-dimensional surface of object in x-axis
Gradient on direction, zy' indicate the gradient of the first side three-dimensional surface of object in the y-axis direction.
Further, S6 includes: to merge to first side 3D shape and second side 3D shape, is obtained
The target three-dimensional image for rebuilding object.
Specifically, the image information of the first side of target is directly acquired by CMOS camera, and passes through above-mentioned calculating and reconstruction
Step can obtain the image for rebuilding the first side of object.Normalized is passed through in the part directly acquired, can reconstruct target
The surface shape S of relative size1。
Then, the image information that target object second side is acquired by CMOS camera and plane mirror, due to passing through
The body surface that CMOS camera is directly acquired and rebuild is relative size, so the mesh got after reflecting mirror reflects
The object surface shape S of relative size also can be obtained in mark information after treatment2.It is continuous, continual due to rebuilding object
Object, so the relative length of object width is certain at tie point.In the present embodiment, S is chosen1Relative size be gauge
It is very little, by S2With S1Zoom factor τ is denoted as in the ratio for rebuilding object junction width.Finally, by S2It is whole multiplied by zoom factor τ,
By S1And S2Two parts splicing, can be obtained the target three-dimensional image after rebuilding.
The present invention can simply and easily carry out target using three-dimensional rebuilding method of the one camera in conjunction with mirror system
Three-dimensional reconstruction, reduces the complexity of system structure, while reducing the error introduced in multiple information measurement.
The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be said that
Specific implementation of the invention is only limited to these instructions.For those of ordinary skill in the art to which the present invention belongs, exist
Under the premise of not departing from present inventive concept, a number of simple deductions or replacements can also be made, all shall be regarded as belonging to of the invention
Protection scope.
Claims (10)
1. a kind of comprehensive polarization three-dimensional rebuilding method characterized by comprising
S1: the luminous intensity of the target polarization image and the first side of rebuilding the first side of object is obtained;
S2: the luminous intensity of the target polarization image and second side of rebuilding object second side opposite with the first side is obtained;
S3: first surface normal is calculated according to the luminous intensity of the target polarization of the first side and the first side, and to the first surface
Normal is corrected;
S4: second surface normal is calculated according to the luminous intensity of the target polarization of second side and second side, and to the second surface
Normal is corrected;
S5: using after correction first surface normal and correction after second surface normal obtain respectively the first side 3D shape and
Second side 3D shape;
S6: merging first side 3D shape and second side 3D shape, obtains the target three for rebuilding object
Tie up image.
2. comprehensive polarization three-dimensional rebuilding method according to claim 1, which is characterized in that the S1 includes:
S11: under light source irradiation, the image data of the first side of object is rebuild in acquisition, obtains the first scene polarization image I1';
S12: three times by the rotation of polarizing film sequence, 45 ° of rotation every time acquires the image data of first side every time after rotation,
The second scene polarization image I is obtained respectively2', third scene polarize I3' and the 4th scene polarization image I4';
S13: to the first scene polarization image I1', the second scene polarization image I2', third scene polarize I3' and the 4th
Scape polarization image I4' be split respectively, obtain first object polarization image I1, the second target polarization image I2, third target it is inclined
Shake image I3With the 4th target polarization image I4;
S14: pass through the second scene polarization image I2' and the 4th scene polarization image I4' obtain the luminous intensity I of first side0。
3. comprehensive polarization three-dimensional rebuilding method according to claim 2, which is characterized in that the S2 includes:
S21: by two reflecting mirrors, the image data of second side of object backlight is rebuild in acquisition, and to the figure of two reflecting mirrors
As data carry out Feature Points Matching, the 5th scene polarization image I of acquisition5';
S22: three times by the rotation of polarizing film sequence, 45 ° of rotation every time acquires the image data of described second side every time after rotation,
The 6th scene polarization image I is obtained respectively6', the 7th scene polarize I7' and the 8th scene polarization image I8';
S23: to the 5th scene polarization image I5', the 6th scene polarization image I6', the 7th scene polarize I7' and the 8th
Scape polarization image I8' be split respectively, obtain the 5th target polarization image I5, the 6th target polarization image I6, the 7th target it is inclined
Shake image I7With the 8th target polarization image I8;
S24: pass through the 6th scene polarization image I6' and the 8th scene polarization image I8' obtain the luminous intensity I of described second side0’。
4. comprehensive polarization three-dimensional rebuilding method according to claim 3, which is characterized in that the S3 includes:
S31: the degree of polarization P of the first side surface is calculated;
S32: the azimuth of the first side surface is calculatedWith the incidence angle θ of the first side surface;
S33: according to the azimuthThe surface normal of first side is calculated with the incidence angle θComponent n in x-axisx
With the component n in y-axisy;
S34: the surface normal is calculatedCorrection coefficient in x-axisWith the correction coefficient in y-axis
S35: pass through correction coefficientWithThe correction component n in x-axis is obtained respectivelykxAnd the correction component n in y-axisky。
5. comprehensive polarization three-dimensional rebuilding method according to claim 4, which is characterized in that the S31 includes:
Utilize formulaCalculate the degree of polarization P of first side surface, wherein I, Q and U indicate to rebuild object the
The Stokes vector of side, its calculation formula is:
Wherein, ExIndicate the vector of electric field in the direction of the x axis, EyIndicate the vector of electric field in the y-axis direction;ILFor circular polarization
Left polarization image;IRFor the right polarization image of circular polarization.
6. comprehensive polarization three-dimensional rebuilding method according to claim 5, which is characterized in that the S32 includes:
S321: first object polarization image I is utilized1, the second target polarization image I2, third target polarization image I3With the 4th mesh
Mark polarization image I4Calculate the azimuth of first side surface
S322: the incidence angle θ of first side surface is calculated according to the degree of polarization P.
7. comprehensive polarization three-dimensional rebuilding method according to claim 6, which is characterized in that the S33 includes:
According to the azimuthThe surface normal of first side is calculated with the incidence angle θComponent n in x-axisxAnd y
Component n on axisy, calculation formula are as follows:
8. comprehensive polarization three-dimensional rebuilding method according to claim 7, which is characterized in that the S34 includes:
S341: being based on Lambertian reflection model, and the simulation normal for rebuilding body surface is established using shape from shading methodIn x-axis
With the analogue component n in y-axissxAnd nsy, calculation formula are as follows:
Wherein, η is body surface albedo, and L is light source direction, LxAnd LyRespectively light source direction is in x-axis and y-axis direction
Component, I0For the luminous intensity of the first side;
S342: the surface normal is calculatedCorrection coefficient in x-axisCalculation formula are as follows:
S343: the surface normal is calculatedCorrection coefficient on the y axisCalculation formula are as follows:
9. comprehensive polarization three-dimensional rebuilding method according to claim 8, which is characterized in that the S35 includes:
S351: according to correction coefficientAnd surface normalComponent n in x-axisxX-axis component n after calculating correctioncx, calculate
Formula are as follows:
S352: according to correction coefficientAnd surface normalComponent n on the y axisyY-axis component n after calculating correctioncy, calculate
Formula are as follows:
10. comprehensive polarization three-dimensional rebuilding method according to claim 9, which is characterized in that the S5 includes:
S51: the x-axis component n after the correction of first surface normal is utilizedcxWith the y-axis component n after correctioncyCalculate the first side three-dimensional shaped
Shape, calculation formula are as follows:
Cos (z)=∫ ∫ ((zx-ncx)2+(zy-ncy)2)dxdy
Wherein, z indicates to rebuild the three-dimensional surface of object, zxIndicate the gradient of the three-dimensional surface of object in the direction of the x axis, zyIt indicates
The gradient of the three-dimensional surface of object in the y-axis direction;
S52: second side 3D shape is calculated using the x-axis component after the correction of second surface normal and the y-axis component after correction.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810787882.9A CN109285213A (en) | 2018-07-18 | 2018-07-18 | Comprehensive polarization three-dimensional rebuilding method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810787882.9A CN109285213A (en) | 2018-07-18 | 2018-07-18 | Comprehensive polarization three-dimensional rebuilding method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109285213A true CN109285213A (en) | 2019-01-29 |
Family
ID=65182981
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810787882.9A Pending CN109285213A (en) | 2018-07-18 | 2018-07-18 | Comprehensive polarization three-dimensional rebuilding method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109285213A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110458960A (en) * | 2019-06-26 | 2019-11-15 | 西安电子科技大学 | A kind of color body three-dimensional rebuilding method based on polarization |
CN111340936A (en) * | 2020-02-17 | 2020-06-26 | 西安电子科技大学 | Passive three-dimensional reconstruction method based on polarization diffuse reflection separation |
WO2021027719A1 (en) * | 2019-08-09 | 2021-02-18 | 南京理工大学 | Reflector-based calibration method for fringe projection system |
CN113706693A (en) * | 2021-08-27 | 2021-11-26 | 武汉大学 | Polarization three-dimensional reconstruction method under low-light-level condition |
CN114731368A (en) * | 2019-12-13 | 2022-07-08 | 索尼集团公司 | Imaging apparatus, information processing apparatus, imaging method, and information processing method |
CN116091361A (en) * | 2023-03-23 | 2023-05-09 | 长春理工大学 | Multi-polarization parameter image fusion method, system and terrain exploration monitor |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2010021972A1 (en) * | 2008-08-18 | 2010-02-25 | Brown University | Surround structured lighting for recovering 3d object shape and appearance |
CN101930628A (en) * | 2010-09-21 | 2010-12-29 | 北京大学 | Monocular-camera and multiplane mirror catadioptric device-based motion capturing method |
CN104224127A (en) * | 2014-09-17 | 2014-12-24 | 西安电子科技大学 | Optical projection tomography device and method based on camera array |
CN104573180A (en) * | 2014-12-02 | 2015-04-29 | 浙江工业大学 | Real-person shoe type copying device and shoe tree manufacturing method based on single-eye multi-angle-of-view robot vision |
CN106556356A (en) * | 2016-12-07 | 2017-04-05 | 西安知象光电科技有限公司 | A kind of multi-angle measuring three-dimensional profile system and measuring method |
CN107680156A (en) * | 2017-09-08 | 2018-02-09 | 西安电子科技大学 | Three-dimensional rebuilding method based on polarization information |
-
2018
- 2018-07-18 CN CN201810787882.9A patent/CN109285213A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2010021972A1 (en) * | 2008-08-18 | 2010-02-25 | Brown University | Surround structured lighting for recovering 3d object shape and appearance |
CN101930628A (en) * | 2010-09-21 | 2010-12-29 | 北京大学 | Monocular-camera and multiplane mirror catadioptric device-based motion capturing method |
CN104224127A (en) * | 2014-09-17 | 2014-12-24 | 西安电子科技大学 | Optical projection tomography device and method based on camera array |
CN104573180A (en) * | 2014-12-02 | 2015-04-29 | 浙江工业大学 | Real-person shoe type copying device and shoe tree manufacturing method based on single-eye multi-angle-of-view robot vision |
CN106556356A (en) * | 2016-12-07 | 2017-04-05 | 西安知象光电科技有限公司 | A kind of multi-angle measuring three-dimensional profile system and measuring method |
CN107680156A (en) * | 2017-09-08 | 2018-02-09 | 西安电子科技大学 | Three-dimensional rebuilding method based on polarization information |
Non-Patent Citations (4)
Title |
---|
KEITH FORBES ET.AL: "Shape-from-Silhouette with Two Mirrors and an Uncalibrated Camera", 《COMPUTER VISION》 * |
方艳霞: "基于激光三角法和光度立体的三维表面重构", 《中国优秀硕士学位论文全文数据库信息科技辑(月刊)》 * |
曾增: "物体全面形三维信息获取技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑(月刊)》 * |
高鹰: "漫反射偏振三维成像关键技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑(月刊)》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110458960A (en) * | 2019-06-26 | 2019-11-15 | 西安电子科技大学 | A kind of color body three-dimensional rebuilding method based on polarization |
WO2021027719A1 (en) * | 2019-08-09 | 2021-02-18 | 南京理工大学 | Reflector-based calibration method for fringe projection system |
US11808564B2 (en) | 2019-08-09 | 2023-11-07 | Nanjing University Of Science And Technology | Calibration method for fringe projection systems based on plane mirrors |
CN114731368A (en) * | 2019-12-13 | 2022-07-08 | 索尼集团公司 | Imaging apparatus, information processing apparatus, imaging method, and information processing method |
CN111340936A (en) * | 2020-02-17 | 2020-06-26 | 西安电子科技大学 | Passive three-dimensional reconstruction method based on polarization diffuse reflection separation |
CN113706693A (en) * | 2021-08-27 | 2021-11-26 | 武汉大学 | Polarization three-dimensional reconstruction method under low-light-level condition |
CN113706693B (en) * | 2021-08-27 | 2023-11-17 | 武汉大学 | Polarization three-dimensional reconstruction method under low-light condition |
CN116091361A (en) * | 2023-03-23 | 2023-05-09 | 长春理工大学 | Multi-polarization parameter image fusion method, system and terrain exploration monitor |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109285213A (en) | Comprehensive polarization three-dimensional rebuilding method | |
Tunwattanapong et al. | Acquiring reflectance and shape from continuous spherical harmonic illumination | |
ES2402229T3 (en) | Image fusion method and device | |
US8576381B2 (en) | Method and apparatus for quantitative 3-D imaging | |
Baek et al. | Image-based acquisition and modeling of polarimetric reflectance. | |
CN104570382B (en) | The device and method of optical image super-resolution are shifted using integral | |
KR20190133229A (en) | Skew Mirror Secondary Imaging | |
KR20150120066A (en) | System for distortion correction and calibration using pattern projection, and method using the same | |
CN111028205B (en) | Eye pupil positioning method and device based on binocular distance measurement | |
RU2734115C2 (en) | Method and device for generating data characterizing a pixel beam | |
Zou et al. | High-accuracy calibration of line-structured light vision sensors using a plane mirror | |
CN115876124A (en) | High-light-reflection surface three-dimensional reconstruction method and device based on polarized structured light camera | |
Wilm et al. | Accurate and simple calibration of DLP projector systems | |
CN109084679B (en) | A kind of 3D measurement and acquisition device based on spatial light modulator | |
Han et al. | Curved LCD based deflectometry method for specular surface measurement | |
US11410378B1 (en) | Image processing for generating three-dimensional shape and spatially-varying reflectance of the object using a deep neural network | |
Zhu et al. | Three-dimensional measurement of fringe projection based on the camera response function of the polarization system | |
Pistellato et al. | A geometric model for polarization imaging on projective cameras | |
Wong et al. | View-synthesized ‘re-calibration’of an array projector for 3D measurement from an arbitrary monocular view | |
Urban et al. | On the Issues of TrueDepth Sensor Data for Computer Vision Tasks Across Different iPad Generations | |
Xu et al. | 3D reconstruction of transparent objects by direct polarized light measurements | |
CN209279884U (en) | Image capture device, 3D information comparison device and mating object generating means | |
Wang et al. | 3D Reconstruction of Weakly Textured Objects Based On Denoising Polarized Images | |
JP2006078291A (en) | Omnidirectional three-dimensional measuring apparatus | |
Xue et al. | Shape Reconstruction of Heritage Assets by Polarization Information |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190129 |
|
RJ01 | Rejection of invention patent application after publication |