CN105954161A - CT-image-based three-dimensional automatic measurement method for particle size of aggregate - Google Patents
CT-image-based three-dimensional automatic measurement method for particle size of aggregate Download PDFInfo
- Publication number
- CN105954161A CN105954161A CN201610195893.9A CN201610195893A CN105954161A CN 105954161 A CN105954161 A CN 105954161A CN 201610195893 A CN201610195893 A CN 201610195893A CN 105954161 A CN105954161 A CN 105954161A
- Authority
- CN
- China
- Prior art keywords
- pixel
- aggregate
- image
- bounding box
- face
- 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
- 239000002245 particle Substances 0.000 title claims abstract description 63
- 238000000691 measurement method Methods 0.000 title abstract 3
- 238000000034 method Methods 0.000 claims abstract description 37
- 238000005259 measurement Methods 0.000 claims abstract description 19
- 239000000203 mixture Substances 0.000 claims abstract description 16
- 238000012360 testing method Methods 0.000 claims abstract description 9
- 239000000463 material Substances 0.000 claims description 89
- 239000010426 asphalt Substances 0.000 claims description 29
- 230000007423 decrease Effects 0.000 claims description 15
- 239000013521 mastic Substances 0.000 claims description 11
- 238000007873 sieving Methods 0.000 claims description 11
- 239000003795 chemical substances by application Substances 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 6
- 230000004069 differentiation Effects 0.000 claims description 6
- 230000005484 gravity Effects 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 claims description 5
- RSWGJHLUYNHPMX-ONCXSQPRSA-N abietic acid Chemical compound C([C@@H]12)CC(C(C)C)=CC1=CC[C@@H]1[C@]2(C)CCC[C@@]1(C)C(O)=O RSWGJHLUYNHPMX-ONCXSQPRSA-N 0.000 claims description 4
- 230000005262 alpha decay Effects 0.000 claims description 3
- 239000004568 cement Substances 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims 1
- 238000004458 analytical method Methods 0.000 abstract description 3
- 238000003041 virtual screening Methods 0.000 abstract description 3
- 150000001875 compounds Chemical class 0.000 description 9
- 238000012216 screening Methods 0.000 description 5
- 238000002591 computed tomography Methods 0.000 description 3
- 239000000853 adhesive Substances 0.000 description 2
- 230000001070 adhesive effect Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 239000002002 slurry Substances 0.000 description 2
- 241001502050 Acis Species 0.000 description 1
- -1 Colophonium Compound Chemical class 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000000627 alternating current impedance spectroscopy Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 239000004567 concrete Substances 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 239000008187 granular material Substances 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/02—Investigating particle size or size distribution
Landscapes
- Chemical & Material Sciences (AREA)
- Dispersion Chemistry (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a CT-image-based three-dimensional automatic measurement method for a particle size of aggregate. The CT-image-based three-dimensional automatic measurement method for a particle size of aggregate is provided for application to tests and numerical analysis of a bituminous mixture. The particle size of each aggregate in the mixture is automatically measured with high precision. According to the method, based on law prediction of pavement performance impacted by a microstructure of the bituminous mixture, and based on a CT image of the bituminous mixture, three-dimensional high-precision automatic measurement of the particle size of aggregate in the mixture is completed to assist in precision three-dimensional virtual screening of aggregate in the mixture.
Description
Technical field
The present invention relates to Asphalt Mixture Experiment and numerical analysis method field, a kind of based on CT image gather materials
Particle diameter three-dimensional method for automatic measurement.
Background technology
Asphalt is to be mutually mixed with asphalt mastic gathering materials of certain grating, be compacted and the composite that formed.
Wherein, mutual embedded squeezing of gathering materials and the skeleton that the constitutes pavement performance important to compound.The difference of grating will cause
The difference of framing structure, thus affect pavement performance.Therefore, it is thus achieved that the accurate grating gathered materials in compound is that research grating is with mixed
Close the important prerequisite of rule between material pavement performance.Realizing this target, key issue accurately measures each aggregate particle
Particle diameter, thus complete the statistics of aggregate grading.
During the mechanical grading gathered materials, owing to aggregate shape is sufficiently complex, it is difficult to ensure the adequacy and thorough of screening
End property, thus cause some aggregate particles not over the minimum sieve aperture of its correspondence, cause bigger grating error.Therefore,
Only developing the accurate measuring technique of single aggregate size, the obtained grating of guarantee has higher precision.
Virtual screening is a kind of based on the digital picture gathered materials or CT profile scanning image analysis calculation aggregate size and unite
The method of meter grating.Such method can be prevented effectively from bigger grating error.The most relatively common virtual method for sieving bag
Include two dimension screening and three-dimensional screening two classes.The mode calculating aggregate size in two class methods is different.The former gathers materials
The digital picture of a certain angle, and then calculate the aggregate size in image or area；The latter uses CT equipment uniformly, equidistantly sweep
Retouching its CT image of acquisition that gathers materials, setting up its three-D profile based on this, and finding this and gather materials and can be obtained by the smallest cross-sectional of sieve aperture
Obtain particle diameter.Obviously, two dimension screening does not consider the 3D shape characteristic gathered materials, it is thus achieved that particle diameter error greatly, with a low credibility；And
The most accurately judge to gather materials in three-dimensional screening the smallest cross-sectional by sieve aperture, could obtain accurate particle diameter.
" Colophonium based on Digital Image Processing mixes volume 32 the 5th phase " Jiaotong University Of East China's journal " journal article in 2015
Close material coarse aggregate method for sieving " utilize digital camera to obtain the image that gathers materials, with the area that gathers materials as key index, enter from two dimension angular
Go the virtual screening gathered materials.But the image that gathers materials of fixed angle cannot reflect the 3D shape characteristic gathered materials, the area gathered materials
Also without direct relation between size and its particle diameter.
Within 2012, volume 42 the 4th phase " Jilin University's journal (engineering version) " journal article is " based on CT tomoscan image
Concrete rough aggregate three-dimensional sieves " according to coarse aggregate CT image, analyze and obtain the discrete profile point of surface of aggregate and utilize
Spherical harmonics method is encrypted, and can algorithm for design find and determine to gather materials and by the control plane of sieve aperture, thus obtain this and gather materials
Particle diameter.But the method asks for controlling plane along the line direction between farthest two the discrete profile point of interior distance of gathering materials, and
The complexity of aggregate shape and physics, geometrical rule determine striked control plane and preferably control interplanar and still suffer from
Certain gap, therefore result in the bigger error of aggregate size result of calculation.
National inventing patent " the thin sight physical model reconstructing method that gathers materials based on bituminous paving test specimen X-ray CT image "
(patent No.: ZL201210172375.7) uses computer graphic image technology, the compound according to compound CT image reconstruction
The three-dimensional geometry physical model of all aggregate particles in test specimen.On this basis, exploitation is based on matrix self adaptation for the present invention
The control plane search technology of three-dimensional rotation, adjusting acquisition matrix can be by the space angle of minimum sieve aperture, at this base
Obtain Optimal Control plane on plinth, finally give high-precision aggregate size.
Summary of the invention
It is an object of the invention to the asphalt for being applied to experimental and numerical analysis and a kind of collection based on CT image is provided
Material particle diameter three-dimensional method for automatic measurement, it is achieved in compound, the high-precision automatic of each aggregate size is measured.
In order to achieve the above object, the technical solution adopted in the present invention is:
Aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: comprise the following steps:
(1) it is, the certain bitumen mixture specimen X-ray of 512 × 512 pixels, sweep spacing with BMP form, resolution
CT gray level image file group is data source, selects an aggregate particle and intercepts its projection segment in each scanogram, protects
Save as gray level image file group Aggregate [n] of BMP form, in order to reconstruct the physical model of this aggregate particle；
(2), travel through gray level image file group Aggregate [n], use contour detecting and modeling algorithm, detect every width
Image Aggregate [i] (i=1,2 ..., n) in the contour pixel of aggregate particle, be stored in contour pixel array Pi(i=1,
2 ..., n), resettle corresponding contoured surface physical model OFi(i=1,2 ..., n), call the bottom modeling function of ACIS7.0
Skinning, it is thus achieved that the aggregate particle physical model M of BODY type；
(3), use 3D minimum bounding box searching algorithm, obtain the 3D minimum bounding box of aggregate particle physical model M
3DMinBoxM；
(4), use based on the aggregate size measuring method controlling plane, according to M and 3DMinBoxMGeometry and
Spatial positional information obtains aggregate size.
Described aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: in step (1), described
Bitumen mixture specimen refers to asphalt material and gathers materials by proper proportion preparation, is cemented to the cylindrical sample of entirety, note half
Footpath is radius, and height is height；
Described X-ray CT gray level image file group refers to use X-ray CT equipment with certain longitudinally spaced scanning Colophonium
Compound test specimen and the profile scanning image array that obtains, every width profile scanning image saves as the gray level image literary composition of BMP form
Part；
Described gray level image file refers to be weighted the color component value of pixel in coloured image according to a certain percentage,
And weighted value is invested each color component and the image file that obtains, wherein, in gray level image, the gray scale of all pixels is all situated between
Between 0 to 255；
Described projection segment refer in a width scanogram surround target gather materials particle image minimum rectangle in segment；
Described physical model refers to the three-dimensional geometrical structure represented with B reps, be designated as model M=(FACE, EDGE,
VERTEX, R), the set in face during wherein FACE represents model, EDGE represents the set on limit, and VERTEX represents the set on summit, R
Represent the syntopy between each element in model；
Record the classification in face: plane, batten face, and the surface equation in face, wherein face is
Element in FACE；
Record the curvilinear equation on limit, the element during wherein edge is EDGE；
Record summit coordinate p (x, y, z), the element during wherein vertex is VERTEX；
R={r1,r2, wherein:
r1={ (face1,face2,edge)|edge∈face1∩face2,edge∈EDGE,facei∈ FACE, i=1,
2}, r1Expression face face1And face2Adjacent to limit edge；
r2={ (edge1,edge2,vertex)|vertex∈edge1∩edge2,vertex∈VERTEX,edgei∈
EDGE, i=1,2}, r2Represent limit edge1With edge2Adjacent to a vertex.
Described aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: in step (2), profile
Detection and modeling algorithm are described as follows step:
1), using the center pixel of image Aggregate [i] and eight neighborhood pixel thereof as ash corresponding to this image
Degree difference Octree GDOiRoot node and the child node of root node, access flag bV=true of these 9 pixels of juxtaposition and root
After the expansible mark bE=true of all child nodes of node, calculate GDOiIn the gray scale difference score value d of all connecting line segmentsFCAnd will
Current level number h is set to 2, and depth capacity H is set to the integer part of smaller's half in Aggregate [i] vertical, horizontal resolution；
2), grey scale difference threshold value automatic parsing algorithm is used to obtain grey scale difference threshold value TG, judge successively each to be positioned at GDOi
In the d of connecting line segment between the node of h layer and its father nodeFCWhether more than TG, the most then the respective pixel of this node is profile
Pixel, is stored in P by this pixeliThe bE=false of this node of juxtaposition, the most only puts the bE=true of this node；
3), traversal GDOiIn the node of h layerIfBE=true, then willThe pixel conduct of bV=false in the eight neighborhood pixel of respective pixelChild node, juxtapositionAll sons
The bE=true of node, the bV=true of child node respective pixel, make h increase 1；
4), continuous service step 2)～step 3) until h=H or GDOiIn pixel corresponding to node on h layer equal
Till contour pixel, complete to build GDOi；
5), according to PiIn each pixel syntopy in image Aggregate [i], to PiIn pixel carry out packet row
Sequence, is arranged pixel the most adjacent by adjacent order and is attributed to one group, if PiOnly comprise one group of pixel, perform step
Rapid 7), step 6 is otherwise performed)；
6), to PiIn two end pixels of each pixel groups, find out another pixel groups that distance is the shortest therewith respectively
End pixels is paired, and to match one of pixel as starting point, another pairing pixel is terminal, with the line between beginning and end
Direction is detection direction, by TGGradually subtract 1 and detect contour pixel, until the contour pixel between beginning and end is at both lines
On direction the most adjacent；
7) the bottom modeling function Spline_Fitting matching P of ACIS7.0, is callediIn pixel, it is thus achieved that FACE type
Contoured surface physical model OFi；
Wherein:
Described eight neighborhood pixel refers to the upper and lower, left and right of a pixel in image, upper left, upper right, lower-left and bottom right eight
Pixel adjacent therewith on individual direction；
Described grey scale difference Octree GDO refers to describe involved pixel during aggregate particle contour pixel detects
Between syntopy and adjacent pixels between the Octree of grey scale difference；
If described Octree refers to be formed by connecting and each knot with straightway according to the filiation between node by passive node
A kind of tree form data structure of point be up to eight child nodes；
Described gray scale difference score value dFCRefer to the gray scale difference between child node and father node respective pixel in GDO；
Described current level number h refers to GDOiCurrent layer number in building process；
Described depth capacity H refers to that built GDO allows the maximum number of plies comprised；
Described contoured surface refers to PiThe fitting result of middle pixel is the plane on border.
Described aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: described contour detecting and
The step 2 of modeling algorithm) in, grey scale difference threshold value automatic parsing algorithm is described as follows step:
1), will gather materials pixel grey scale range limitIt is entered as in grey level histogram corresponding to Aggregate [i]
High-gray level value；
2), variable t is entered as
3), allow t successively decrease 1 every time, untilMiddle pixel accounts for the ratio of whole pixels in Aggregate [i] and exceedes collection
Material pixel ratio pAG；
4), allow t successively decrease 1 every time, till no longer comprising, in the pixel that gray scale is t, pixel of gathering materials, pixel of gathering materials ash
Degree range lower limitIt is entered as t；
5), allow t successively decrease 1, till no longer comprising asphalt mastic pixel in the pixel that gray scale is t, by asphalt adhesive every time
Slurry pixel grey scale range limitIt is entered as t；
6), by grey scale difference threshold value TGIt is entered asWithDifference；
Wherein:
Described pixel grey scale range limit of gathering materialsRefer to gather materials in grey level histogram corresponding to Aggregate [i] pixel
The upper limit of tonal range；
Described grey level histogram refers in respective image the cartogram of pixel quantity in each gray level；
Described pixel ratio p of gathering materialsAGWhat the pixel that refers to gather materials accounted for Aggregate [i] all pixels estimates ratio；
The described pixel grey scale range lower limit that gathers materialsRefer to gather materials in grey level histogram corresponding to Aggregate [i] pixel
The lower limit of tonal range；
Described asphalt mastic pixel grey scale range limitRefer to grey level histogram medium pitch corresponding to Aggregate [i]
The upper limit of rubber cement pixel grey scale scope；
Described grey scale difference threshold value TGRefer to judge that two any pixels are the most respectively gathered materials and asphalt mastic pixel
Grey scale difference minima.
Described aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: in described step (3),
3D minimum bounding box searching algorithm is described as follows step:
1) the bottom modeling function Mass_Proporties, using ACIS7.0 calculates the weight of aggregate particle physical model M
Heart GPM, M is moved to GPMThe position overlapped with global coordinate system initial point, making rotary shaft is x-axis；
2) the bottom modeling function Get_Bounding_Box, using ACIS7.0 obtains the current 3D bounding box of M
3DBeBoxMAnd calculate its volume BeVMBoxAfter, with GPMCentered by, make M rotate forward θ=1 ° around rotary shaft, it is thus achieved that after M rotates
3D bounding box 3DAfBoxMAnd calculate its volume AfVMBoxIf, BeVMBox>AfVMBox, by the orientation before M rollback to rotation, and
Set direction of rotation as rotary shaft forward, otherwise make M around rotary shaft counter-rotating θ=1 °, if now BeVMBox>AfVMBox, by M
Direction of rotation is set reverse as rotary shaft behind orientation before rollback extremely rotation, if now BeVMBox<AfVMBox, rotation terminates, and performs
Step 5)；
3), with GPMCentered by, along step 2) direction of rotation that sets, determine actual θ value according to 3D corner decay rule
After, M is rotated θ angle；
4), judge whether 3D terminating rotation condition is reached, if so, rotate and terminate, perform step 5), otherwise perform step
3)；
5) making rotary shaft, successively is y-axis and z-axis, performs step 2)～4), terminate until M all rotates around three coordinate axess,
Obtain the 3D minimum bounding box 3DMinBox of MM；
Wherein:
That described 3D bounding box refers to be surrounded aggregate particle physical model M completely, each limit all with a certain coordinate axes
Parallel minimum cuboid；
Described 3D corner decay rule refers to differentiate that M is at the 3D not rotated, rotate under θ angle and three kinds of situations of θ+1 angle
Bounding box volume successively decreases the most successively, if so, determines that the anglec of rotation is θ, if it is not, again differentiate after then being halved by θ, until obtaining
Differentiation result certainly or θ decay to 1 °, and wherein the initial value of θ is set to 8 °；
Described 3D terminating rotation condition refer to M not rotating, in the 3D bounding box volume that rotates under 1 ° of two kinds of situation, the latter
Bigger；
Described 3D minimum bounding box 3DMinBoxMRefer to that M carries out any direction, arbitrarily angled with any center around any axle
Rotation after volume reckling in obtained 3D bounding box.
Described aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: in described step (4),
It is described as follows step based on the aggregate size measuring method controlling plane:
1) the 3D minimum bounding box 3DMinBox of aggregate particle physical model M, is obtainedMIn longest edge one, it is judged that and
Determine coordinate axes Axis parallel with itL；
2), use ACIS7.0 bottom modeling function Section, with 3DMinBoxMAxis minimum in 8 summitsLSit
Scale value is AxisLOn original position, maximum AxisLCoordinate figure is AxisLOn final position, along and AxisLVertical side
To, to be spaced intervalMM is carried out cutting, it is thus achieved that q part Seck(k=1,2 ..., q), and use the bottom of ACIS7.0
Modeling function Mass_Proporties calculates each SeckCenter of gravity
3), calculating is derived fromExtremelyVectorial Veck,k+1(k=1,2 ..., q-1), use the end of ACIS7.0
Layer modeling function Section, respectively edge and vector Vec(k,k+1)(k=1,2 ..., q-1) vertical direction, mistakeIntercept the cross section of M, obtain the set SFS={SF of FACE style cross-sectionsk| k=1,2 ..., 2q-2}；
4) the bottom modeling function Get_Face_Normal, using ACIS7.0 obtains SFS middle section SFk(k=1,
2 ..., 2q-2) normal vector, it is judged that it is with coordinate axes AxisLAngle between forward, obtains SFkSummit and with it be
The heart, rotates SFkTo its normal vector and AxisLParallel orientation；
5) the bottom modeling function Get_Face_Box, using ACIS7.0 obtains SFkCurrent 2D bounding boxCalculate and obtain its centerAnd areaAfter, by SFkMove toWith global coordinate system initial point
The position overlapped, withCentered by, make SFkAround AxisLRotate forward α=1 °, it is thus achieved that SFkPostrotational 2D bounding boxAnd calculate its areaIfBy SFkOrientation before rollback extremely rotation, and set
Determining direction of rotation is AxisLForward, otherwise by SFkSF is made behind orientation before rollback extremely rotationkAround AxisLCounter-rotating α=1 °,
If nowThen by SFkDirection of rotation is set as Axis behind orientation before rollback extremely rotationLInversely, if this
TimeRotation terminates, by SFkCurrent 2D bounding box as its 2D minimum bounding boxHold
Row step 8)；
6), withCentered by, along step 5) direction of rotation that sets, determine actual α value according to 2D corner decay rule
After, by SFkRotation alpha angle；
7), judge whether 2D terminating rotation condition is reached, if so, by SFkCurrent 2D bounding box as its 2D parcel
Enclose boxOtherwise perform step 6)；
8), to each cross section SF in SFSkPerform step 5 successively)～7), until obtaining setThe relatively longer sides length of side of each element in set, by unit the longest for the longer sides length of side
Cross section corresponding to element is as controlling plane CF, and the corresponding longer sides length of side is sieved size as the minimum gathered materials, with closest
And the standard screen screen size being not less than minimum size of sieving is aggregate size；
Wherein:
Described cross section SFkNormal vector refer to and cross section SFkVertical vector；
It is equal that described 2D bounding box refers to limit that the plane that parallel with a certain coordinate surface be surrounded completely, each
The minimum rectangle parallel with a certain coordinate axes；
Described 2D corner decay rule refers to differentiate SFkNot rotating, under rotation alpha angle and three kinds of situations of α+1 angle
2D bounding box area successively decreases the most successively, if so, determines that the anglec of rotation is α, if it is not, again differentiate after then being halved by α, until obtaining
The differentiation result that must affirm or α decay to 1 °, and wherein the initial value of α is set to 8 °；
Described 2D terminating rotation condition refers to SFkNot rotating, in the 2D bounding box area that rotates under 1 ° of two kinds of situation, after
Person is bigger；
Described 2D minimum bounding boxRefer to SFkThis center and SF is walked around with any centerkVertical rotation
Rotating shaft carries out any direction, rotate at any angle after area reckling in obtained 2D bounding box；
Described control plane CF refers to determine to gather materials particle entities model M can be by the cross section of minimum sieve aperture；
Described minimum size of sieving refers to the minimum sieve hole dimension that can pass through that gathers materials；
Described aggregate size refers to gathering materials the minimum sieve hole dimension that can pass through in standard screen；
Described standard screen refers to be made up of the bushing screen in one group of multiple different size aperture, for sieving what different size gathered materials
Instrument, screen size is followed successively by from small to large: 1.18mm, 2.36mm, 4.75mm, 9.5mm, 12.5mm, 19mm.
Described aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: described ACIS7.0 be by
The THREE DIMENSION GEOMETRIC MODELING engine based on Object Oriented technology that Spatial company of the U.S. produces.
Beneficial effects of the present invention is as follows:
The affecting laws of its pavement performance is predicted by the present invention based on asphalt microscopical structure, according to asphalt
CT image, completes the three-dimensional high-precision of aggregate size in compound and automatically measures, with auxiliary realize compound gathers materials accurate
Three-dimensional sieves.
Accompanying drawing explanation
Fig. 1 a is certain bitumen mixture specimen X-ray CT scan image sequence.
Fig. 1 b is the single aggregate particle chosen projection segment in test specimen X-ray CT scan image sequence.
Fig. 2 a is aggregate particle testing result figure of contour pixel in each projection segment.
Fig. 2 b is the contoured surface physical model that aggregate particle is corresponding in each projection segment.
Fig. 3 is aggregate particle physical model.
Fig. 4 a is the 3D bounding box before aggregate particle physical model rotates.
Fig. 4 b is aggregate particle physical model postrotational 3D minimum bounding box.
Fig. 4 c is the control plane of aggregate particle physical model.
Fig. 4 d is the 2D minimum bounding box that aggregate particle physical model controls plane.
Detailed description of the invention
Aggregate size three-dimensional method for automatic measurement based on CT image, comprises the following steps:
(1) it is, the certain bitumen mixture specimen X-ray of 512 × 512 pixels, sweep spacing with BMP form, resolution
CT gray level image file group is data source, selects an aggregate particle and intercepts its projection segment in each scanogram, protects
Save as gray level image file group Aggregate [n] of BMP form, in order to reconstruct the physical model of this aggregate particle, wherein:
Bitumen mixture specimen refers to asphalt material and gathers materials by proper proportion preparation, is cemented to the cylindrical sample of entirety
This, note radius is radius, and height is height；
X-ray CT gray level image file group refers to use X-ray CT equipment to mix with certain longitudinally spaced scanning Colophonium
Material test specimen and the profile scanning image array that obtains, every width profile scanning image saves as the gray level image file of BMP form；
Gray level image file refers to be weighted the color component value of pixel in coloured image according to a certain percentage, and will
The image file that weighted value invests each color component and obtains, wherein, in gray level image, the gray scale of all pixels all arrives between 0
Between 255；
Projection segment refer in a width scanogram surround target gather materials particle image minimum rectangle in segment；
Physical model refers to the three-dimensional geometrical structure represented with B reps, be designated as model M=(FACE, EDGE,
VERTEX, R), the set in face during wherein FACE represents model, EDGE represents the set on limit, and VERTEX represents the set on summit, R
Represent the syntopy between each element in model；
Record the classification in face: plane, batten face, and the surface equation in face, wherein face is
Element in FACE；
Record the curvilinear equation on limit, the element during wherein edge is EDGE；
Record summit coordinate p (x, y, z), the element during wherein vertex is VERTEX；
R={r1,r2, wherein:
r1={ (face1,face2,edge)|edge∈face1∩face2,edge∈EDGE,facei∈ FACE, i=1,
2}, r1Expression face face1And face2Adjacent to limit edge；
r2={ (edge1,edge2,vertex)|vertex∈edge1∩edge2,vertex∈VERTEX,edgei∈
EDGE, i=1,2}, r2Represent limit edge1With edge2Adjacent to a vertex；
(2), travel through gray level image file group Aggregate [n], use contour detecting and modeling algorithm, detect every width
Image Aggregate [i] (i=1,2 ..., n) in the contour pixel of aggregate particle, be stored in contour pixel array Pi(i=1,
2 ..., n), resettle corresponding contoured surface physical model OFi(i=1,2 ..., n), call the bottom modeling function of ACIS7.0
Skinning, it is thus achieved that the aggregate particle physical model M of BODY type；
(3), use 3D minimum bounding box searching algorithm, obtain the 3D minimum bounding box of aggregate particle physical model M
3DMinBoxM；
(4), use based on the aggregate size measuring method controlling plane, according to M and 3DMinBoxMGeometry and sky
Between positional information obtain aggregate size.
In step (2), contour detecting and modeling algorithm are described as follows step:
1), using the center pixel of image Aggregate [i] and eight neighborhood pixel thereof as gray scale corresponding to this image
Difference Octree GDOiRoot node and the child node of root node, access flag bV=true of these 9 pixels of juxtaposition and root knot
After the expansible mark bE=true of all child nodes of point, calculate GDOiIn the gray scale difference score value d of all connecting line segmentsFCAnd ought
Front level number h is set to 2, and depth capacity H is set to the integer part of smaller's half in Aggregate [i] vertical, horizontal resolution；
2), grey scale difference threshold value automatic parsing algorithm is used to obtain grey scale difference threshold value TG, judge successively each to be positioned at GDOi
In the d of connecting line segment between the node of h layer and its father nodeFCWhether more than TG, the most then the respective pixel of this node is profile
Pixel, is stored in P by this pixeliThe bE=false of this node of juxtaposition, the most only puts the bE=true of this node；
3), traversal GDOiIn the node of h layerIfBE=true, then willThe pixel conduct of bV=false in the eight neighborhood pixel of respective pixelChild node, juxtapositionAll sons
The bE=true of node, the bV=true of child node respective pixel, make h increase 1；
4), continuous service step 2)～step 3) until h=H or GDOiIn pixel corresponding to node on h layer equal
Till contour pixel, complete to build GDOi；
5), according to PiIn each pixel syntopy in image Aggregate [i], to PiIn pixel carry out packet row
Sequence, is arranged pixel the most adjacent by adjacent order and is attributed to one group, if PiOnly comprise one group of pixel, perform step
Rapid 7), step 6 is otherwise performed)；
6), to PiIn two end pixels of each pixel groups, find out another pixel groups that distance is the shortest therewith respectively
End pixels is paired, and to match one of pixel as starting point, another pairing pixel is terminal, with the line between beginning and end
Direction is detection direction, by TGGradually subtract 1 and detect contour pixel, until the contour pixel between beginning and end is at both lines
On direction the most adjacent；
7) the bottom modeling function Spline_Fitting matching P of ACIS7.0, is callediIn pixel, it is thus achieved that FACE type
Contoured surface physical model OFi；
Wherein:
Eight neighborhood pixel refers to the upper and lower, left and right of a pixel in image, upper left, upper right, lower-left and eight, bottom right side
Pixel the most adjacent；
Grey scale difference Octree GDO is during referring to describe the detection of aggregate particle contour pixel between involved pixel
The Octree of grey scale difference between syntopy and adjacent pixels；
If Octree refers to be formed by connecting with straightway according to the filiation between node by passive node and each node
There is a kind of tree form data structure of eight child nodes more；
Gray scale difference score value dFCRefer to the gray scale difference between child node and father node respective pixel in GDO；
Current level number h refers to GDOiCurrent layer number in building process；
Depth capacity H refers to that built GDO allows the maximum number of plies comprised；
Contoured surface refers to PiThe fitting result of middle pixel is the plane on border.
Contour detecting and the step 2 of modeling algorithm) in, grey scale difference threshold value automatic parsing algorithm is described as follows step:
1), will gather materials pixel grey scale range limitIt is entered as in grey level histogram corresponding to Aggregate [i]
High-gray level value；
2), variable t is entered as
3), allow t successively decrease 1 every time, untilMiddle pixel accounts for the ratio of whole pixels in Aggregate [i] and exceedes collection
Material pixel ratio pAG；
4), allow t successively decrease 1 every time, till no longer comprising, in the pixel that gray scale is t, pixel of gathering materials, pixel of gathering materials ash
Degree range lower limitIt is entered as t；
5), allow t successively decrease 1, till no longer comprising asphalt mastic pixel in the pixel that gray scale is t, by asphalt adhesive every time
Slurry pixel grey scale range limitIt is entered as t；
6), by grey scale difference threshold value TGIt is entered asWithDifference；
Wherein:
Gather materials pixel grey scale range limitRefer to gather materials in grey level histogram corresponding to Aggregate [i] pixel grey scale
The upper limit of scope；
Grey level histogram refers in respective image the cartogram of pixel quantity in each gray level；
Gather materials pixel ratio pAGWhat the pixel that refers to gather materials accounted for Aggregate [i] all pixels estimates ratio；
Gather materials pixel grey scale range lower limitRefer to gather materials in grey level histogram corresponding to Aggregate [i] pixel grey scale
The lower limit of scope；
Asphalt mastic pixel grey scale range limitRefer to grey level histogram medium pitch rubber cement corresponding to Aggregate [i]
The upper limit of pixel grey scale scope；
Grey scale difference threshold value TGRefer to judge that two any pixels are the most respectively gathered materials and the ash of asphalt mastic pixel
Degree difference minima.
In step (3), 3D minimum bounding box searching algorithm is described as follows step:
1) the bottom modeling function Mass_Proporties, using ACIS7.0 calculates the weight of aggregate particle physical model M
Heart GPM, M is moved to GPMThe position overlapped with global coordinate system initial point, making rotary shaft is x-axis；
2) the bottom modeling function Get_Bounding_Box, using ACIS7.0 obtains the current 3D bounding box of M
3DBeBoxMAnd calculate its volume BeVMBoxAfter, with GPMCentered by, make M rotate forward θ=1 ° around rotary shaft, it is thus achieved that after M rotates
3D bounding box 3DAfBoxMAnd calculate its volume AfVMBoxIf, BeVMBox>AfVMBox, by the orientation before M rollback to rotation, and
Set direction of rotation as rotary shaft forward, otherwise make M around rotary shaft counter-rotating θ=1 °, if now BeVMBox>AfVMBox, by M
Direction of rotation is set reverse as rotary shaft behind orientation before rollback extremely rotation, if now BeVMBox<AfVMBox, rotation terminates, and performs
Step 5)；
3), with GPMCentered by, along step 2) direction of rotation that sets, determine actual θ value according to 3D corner decay rule
After, M is rotated θ angle；
4), judge whether 3D terminating rotation condition is reached, if so, rotate and terminate, perform step 5), otherwise perform step
3)；
5) making rotary shaft, successively is y-axis and z-axis, performs step 2)～4), terminate until M all rotates around three coordinate axess,
Obtain the 3D minimum bounding box 3DMinBox of MM；
Wherein:
That 3D bounding box refers to be surrounded aggregate particle physical model M completely, each limit is all parallel with a certain coordinate axes
Minimum cuboid；
3D corner decay rule refers to differentiate that M surrounds at the 3D not rotated, rotate under θ angle and three kinds of situations of θ+1 angle
Box body is long-pending to successively decrease the most successively, if so, determines that the anglec of rotation is θ, if it is not, again differentiate after then being halved by θ, until obtaining certainly
Differentiation result or θ decay to 1 °, wherein the initial value of θ is set to 8 °；
3D terminating rotation condition refer to M not rotating, in the 3D bounding box volume that rotates under 1 ° of two kinds of situation, the latter is relatively
Greatly；
3D minimum bounding box 3DMinBoxMRefer to that M carries out any direction with any center around any axle, revolves at any angle
Volume reckling in obtained 3D bounding box after turning.
In step (4), it is described as follows step based on the aggregate size measuring method controlling plane:
1) the 3D minimum bounding box 3DMinBox of aggregate particle physical model M, is obtainedMIn longest edge one, it is judged that and
Determine coordinate axes Axis parallel with itL；
2), use ACIS7.0 bottom modeling function Section, with 3DMinBoxMAxis minimum in 8 summitsLSit
Scale value is AxisLOn original position, maximum AxisLCoordinate figure is AxisLOn final position, along and AxisLVertical side
To, to be spaced intervalMM is carried out cutting, it is thus achieved that q part Seck(k=1,2 ..., q), and use the bottom of ACIS7.0
Modeling function Mass_Proporties calculates each SeckCenter of gravity
3), calculating is derived fromExtremelyVectorial Veck,k+1(k=1,2 ..., q-1), use the end of ACIS7.0
Layer modeling function Section, respectively edge and vector Vec(k,k+1)(k=1,2 ..., q-1) vertical direction, mistakeIntercept the cross section of M, obtain the set SFS={SF of FACE style cross-sectionsk| k=1,2 ..., 2q-2}；
4) the bottom modeling function Get_Face_Normal, using ACIS7.0 obtains SFS middle section SFk(k=1,
2 ..., 2q-2) normal vector, it is judged that it is with coordinate axes AxisLAngle between forward, obtains SFkSummit and with it be
The heart, rotates SFkTo its normal vector and AxisLParallel orientation；
5) the bottom modeling function Get_Face_Box, using ACIS7.0 obtains SFkCurrent 2D bounding boxCalculate and obtain its centerAnd areaAfter, by SFkMove toWith global coordinate system initial point
The position overlapped, withCentered by, make SFkAround AxisLRotate forward α=1 °, it is thus achieved that SFkPostrotational 2D bounding boxAnd calculate its areaIfBy SFkOrientation before rollback extremely rotation, and set
Determining direction of rotation is AxisLForward, otherwise by SFkSF is made behind orientation before rollback extremely rotationkAround AxisLCounter-rotating α=1 °,
If nowThen by SFkDirection of rotation is set as Axis behind orientation before rollback extremely rotationLInversely, if this
TimeRotation terminates, by SFkCurrent 2D bounding box as its 2D minimum bounding box
Perform step 8)；
6), withCentered by, along step 5) direction of rotation that sets, determine actual α value according to 2D corner decay rule
After, by SFkRotation alpha angle；
7), judge whether 2D terminating rotation condition is reached, if so, by SFkCurrent 2D bounding box as its 2D parcel
Enclose boxOtherwise perform step 6)；
8), to each cross section SF in SFSkPerform step 5 successively)～7), until obtaining setThe relatively longer sides length of side of each element in set, by unit the longest for the longer sides length of side
Cross section corresponding to element is as controlling plane CF, and the corresponding longer sides length of side is sieved size as the minimum gathered materials, with closest
And the standard screen screen size being not less than minimum size of sieving is aggregate size；
Wherein:
Cross section SFkNormal vector refer to and cross section SFkVertical vector；
2D bounding box refer to limit that the plane that parallel with a certain coordinate surface is surrounded completely, each all and certain
The minimum rectangle that one coordinate axes is parallel；
2D corner decay rule refers to differentiate SFk2D bag not rotating, under rotation alpha angle and three kinds of situations of α+1 angle
Enclosing box area to successively decrease the most successively, if so, determine that the anglec of rotation is α, if it is not, again differentiate after then α being halved, agreeing until obtaining
Fixed differentiation result or α decay to 1 °, and wherein the initial value of α is set to 8 °；
2D terminating rotation condition refers to SFkNot rotating, in the 2D bounding box area that rotates under 1 ° of two kinds of situation, the latter is relatively
Greatly；
2D minimum bounding boxRefer to SFkThis center and SF is walked around with any centerkVertical rotary shaft
Carry out any direction, rotate at any angle after area reckling in obtained 2D bounding box；
Control plane CF refers to determine to gather materials particle entities model M can be by the cross section of minimum sieve aperture；
Minimum size of sieving refers to the minimum sieve hole dimension that can pass through that gathers materials；
Aggregate size refers to gathering materials the minimum sieve hole dimension that can pass through in standard screen；
Standard screen refers to be made up of the bushing screen in one group of multiple different size aperture, for sieving the work that different size gathers materials
Tool, screen size is followed successively by from small to large: 1.18mm, 2.36mm, 4.75mm, 9.5mm, 12.5mm, 19mm.
In the present invention, with C Plus Plus, based on ACIS kernel, it is achieved that algorithm described in the invention, and with certain Colophonium
The X-ray CT scan image of compound test specimen is data source, completes the three-dimensional of selected aggregate size in compound and automatically surveys
Amount.
(1) with a diameter of 150mm, highly the cylindrical bitumen mixture specimen scanning step for 164mm is the X-of 1mm
Ray CT profile scanning image file group is data source, selects an aggregate particle as shown in Figure 1a and intercepts as shown in Figure 1 b
This aggregate particle projection segment in each scanogram, save as Aggregate [4], in order to reconstruct this aggregate particle
Physical model；
(2) traversal gray level image file group Aggregate [4], for each image, sets pixel ratio p of gathering materialsAG,
Big depth H, and use grey scale difference threshold value automatic parsing algorithm to obtain grey scale difference threshold value TG, wherein pAG、TGValue such as table with H
Shown in 1, use contour detecting and modeling algorithm, detect in Aggregate [i] (i=1,2,3,4) as shown in Figure 2 a and gather materials
The contour pixel of granule, is stored in contour pixel array Pi(i=1,2,3,4), resettles P as shown in Figure 2 biMiddle pixel is corresponding
Contoured surface physical model OFi(i=1,2,3,4), calls the bottom modeling function Skinning of ACIS7.0, it is thus achieved that as shown in Figure 3
The aggregate particle physical model M of BODY type；
Pixel ratio p of gathering materials that table 1 each gray level image file is correspondingAG, depth capacity H and grey scale difference threshold value TG
| Gray level image file Aggregate [i] | Gather materials pixel ratio pAG | Depth capacity H | Grey scale difference threshold value TG |
| I=1 | 0.9 | 17 | 12 |
| I=2 | 0.9 | 19 | 11 |
| I=3 | 0.9 | 12 | 10 |
| I=4 | 0.9 | 11 | 11 |
(3) obtain the bounding box before aggregate particle physical model M rotates, as shown in fig. 4 a, use 3D minimum bounding box to search
Rope algorithm, obtains the 3D minimum bounding box 3DMinBox of aggregate particle physical model MM, as shown in Figure 4 b, wherein the center of gravity of M is
(217.48,139.86,538.74), it is as shown in table 2 rotating around direction of rotation and the anglec of rotation of x-axis, y-axis and z-axis.
Direction that table 2 M rotates rotating around each coordinate axes and angle
| Coordinate axes | X-axis | Y-axis | Z-axis |
| Direction of rotation and angle | 0 | 0 | Reverse 7 ° |
(4) understand and 3DMinBoxMThe coordinate axes Axis that longest edge is parallelLFor x-axis, set intervalM=6.66, make
By aggregate size measuring method based on control plane, M is divided into 10 parts Seck(k=1,2 ..., 10), the most often
The center of gravity of individual partAs shown in table 3, mistakeAlong vector Veck,k+1(k=1,2 ..., 9) direction cuts
Take the cross section of M, it is thus achieved that cross section set SFS={SFk| k=1,2 ..., 18}, wherein, control plane CF and minimum 2D bounding box thereof
Respectively as shown in Fig. 4 c and Fig. 4 d, finally obtain the minimum gathered materials and sieve a size of 48.36 length in pixels, i.e. 14.52mm, thus
Aggregate size is 19mm.
3 10 Sec of tablekBarycentric coodinates
Claims (7)
1. aggregate size three-dimensional method for automatic measurement based on CT image, it is characterised in that: comprise the following steps:
(1) it is, the certain bitumen mixture specimen X-ray CT ash of 512 × 512 pixels, sweep spacing with BMP form, resolution
Degree image file group is data source, selects an aggregate particle and intercepts its projection segment in each scanogram, saving as
Gray level image file group Aggregate [n] of BMP form, in order to reconstruct the physical model of this aggregate particle；
(2), travel through gray level image file group Aggregate [n], use contour detecting and modeling algorithm, detect each image
Aggregate [i] (i=1,2 ..., n) in the contour pixel of aggregate particle, be stored in contour pixel array Pi(i=1,2 ...,
N), corresponding contoured surface physical model OF is resettledi(i=1,2 ..., n), call the bottom modeling function of ACIS7.0
Skinning, it is thus achieved that the aggregate particle physical model M of BODY type；
(3), use 3D minimum bounding box searching algorithm, obtain the 3D minimum bounding box 3DMinBox of aggregate particle physical model MM；
(4), use based on the aggregate size measuring method controlling plane, according to M and 3DMinBoxMGeometry and space bit
Put acquisition of information aggregate size.
Aggregate size three-dimensional method for automatic measurement based on CT image the most according to claim 1, it is characterised in that: step
(1), in, described bitumen mixture specimen refers to asphalt material and gathers materials by proper proportion preparation, is cemented to the cylinder of entirety
Sample, note radius is radius, and height is height；
Described X-ray CT gray level image file group refers to use X-ray CT equipment to mix with certain longitudinally spaced scanning Colophonium
Material test specimen and the profile scanning image array that obtains, every width profile scanning image saves as the gray level image file of BMP form；
Described gray level image file refers to be weighted the color component value of pixel in coloured image according to a certain percentage, and will
The image file that weighted value invests each color component and obtains, wherein, in gray level image, the gray scale of all pixels all arrives between 0
Between 255；
Described projection segment refer in a width scanogram surround target gather materials particle image minimum rectangle in segment；
Described physical model refers to the three-dimensional geometrical structure represented with B reps, be designated as model M=(FACE, EDGE,
VERTEX, R), the set in face during wherein FACE represents model, EDGE represents the set on limit, and VERTEX represents the set on summit, R
Represent the syntopy between each element in model；
Record the classification in face: plane, batten face, and the surface equation in face, during wherein face is FACE
Element；
Record the curvilinear equation on limit, the element during wherein edge is EDGE；
Record summit coordinate p (x, y, z), the element during wherein vertex is VERTEX；
R={r1,r2, wherein:
r1={ (face1,face2,edge)|edge∈face1∩face2,edge∈EDGE,facei∈ FACE, i=1,2}, r1
Expression face face1And face2Adjacent to limit edge；
r2={ (edge1,edge2,vertex)|vertex∈edge1∩edge2,vertex∈VERTEX,edgei∈EDGE,i
=1,2}, r2Represent limit edge1With edge2Adjacent to a vertex.
Aggregate size three-dimensional method for automatic measurement based on CT image the most according to claim 1, it is characterised in that: step
(2), in, contour detecting and modeling algorithm are described as follows step:
1), using the center pixel of image Aggregate [i] and eight neighborhood pixel thereof as grey scale difference corresponding to this image
Octree GDOiRoot node and the child node of root node, access flag bV=true of these 9 pixels of juxtaposition and root node institute
After having the expansible mark bE=true of child node, calculate GDOiIn the gray scale difference score value d of all connecting line segmentsFCAnd by current layer
Number h is set to 2, and depth capacity H is set to the integer part of smaller's half in Aggregate [i] vertical, horizontal resolution；
2), grey scale difference threshold value automatic parsing algorithm is used to obtain grey scale difference threshold value TG, judge successively each to be positioned at GDOiIn
The d of connecting line segment between the node of h layer and its father nodeFCWhether more than TG, the most then the respective pixel of this node is wire-frame image
Element, is stored in P by this pixeliThe bE=false of this node of juxtaposition, the most only puts the bE=true of this node；
3), traversal GDOiIn the node of h layerIfBE=true, then will
The pixel conduct of bV=false in the eight neighborhood pixel of respective pixelChild node, juxtapositionAll child nodes
BE=true, the bV=true of child node respective pixel, make h increase 1；
4), continuous service step 2)～step 3) until h=H or GDOiIn pixel corresponding to node on h layer be profile
Till pixel, complete to build GDOi；
5), according to PiIn each pixel syntopy in image Aggregate [i], to PiIn pixel carry out packet sequencing,
Pixel the most adjacent is arranged by adjacent order and is attributed to one group, if PiOnly comprise one group of pixel, perform step
7), step 6 is otherwise performed)；
6), to PiIn two end pixels of each pixel groups, find out the end picture of another pixel groups that distance is the shortest therewith respectively
Element is paired, and to match one of pixel as starting point, another pairing pixel is terminal, with the line direction between beginning and end is
Detection direction, by TGGradually subtract 1 and detect contour pixel, until the contour pixel between beginning and end is on both line directions
The most adjacent；
7) the bottom modeling function Spline_Fitting matching P of ACIS7.0, is callediIn pixel, it is thus achieved that the wheel of FACE type
Profile surface physical model OFi；
Wherein:
Described eight neighborhood pixel refers to the upper and lower, left and right of a pixel in image, upper left, upper right, lower-left and eight, bottom right side
Pixel the most adjacent；
Described grey scale difference Octree GDO is during referring to describe the detection of aggregate particle contour pixel between involved pixel
The Octree of grey scale difference between syntopy and adjacent pixels；
If described Octree refers to be formed by connecting with straightway according to the filiation between node by passive node and each node
There is a kind of tree form data structure of eight child nodes more；
Described gray scale difference score value dFCRefer to the gray scale difference between child node and father node respective pixel in GDO；
Described current level number h refers to GDOiCurrent layer number in building process；
Described depth capacity H refers to that built GDO allows the maximum number of plies comprised；
Described contoured surface refers to PiThe fitting result of middle pixel is the plane on border.
Aggregate size three-dimensional method for automatic measurement based on CT image the most according to claim 3, it is characterised in that: described
Contour detecting and the step 2 of modeling algorithm) in, grey scale difference threshold value automatic parsing algorithm is described as follows step:
1), will gather materials pixel grey scale range limitThe maximum ash being entered as in grey level histogram corresponding to Aggregate [i]
Angle value；
2), variable t is entered as
3), allow t successively decrease 1 every time, untilMiddle pixel accounts for the ratio of whole pixels in Aggregate [i] and exceedes the picture that gathers materials
Element ratio pAG；
4), allowing t successively decrease 1, till no longer comprising, in the pixel that gray scale is t, pixel of gathering materials, will gather materials pixel grey scale model every time
Enclose lower limitIt is entered as t；
5), allow t successively decrease 1, till no longer comprising asphalt mastic pixel in the pixel that gray scale is t, by asphalt mastic picture every time
The element tonal range upper limitIt is entered as t；
6), by grey scale difference threshold value TGIt is entered asWithDifference；
Wherein:
Described pixel grey scale range limit of gathering materialsRefer to gather materials in grey level histogram corresponding to Aggregate [i] pixel grey scale
The upper limit of scope；
Described grey level histogram refers in respective image the cartogram of pixel quantity in each gray level；
Described pixel ratio p of gathering materialsAGWhat the pixel that refers to gather materials accounted for Aggregate [i] all pixels estimates ratio；
The described pixel grey scale range lower limit that gathers materialsRefer to gather materials in grey level histogram corresponding to Aggregate [i] pixel grey scale
The lower limit of scope；
Described asphalt mastic pixel grey scale range limitRefer to grey level histogram medium pitch rubber cement corresponding to Aggregate [i]
The upper limit of pixel grey scale scope；
Described grey scale difference threshold value TGRefer to judge that two any pixels are the most respectively gathered materials and the gray scale of asphalt mastic pixel
Difference minima.
Aggregate size three-dimensional method for automatic measurement based on CT image the most according to claim 1, it is characterised in that: described
In step (3), 3D minimum bounding box searching algorithm is described as follows step:
1) the bottom modeling function Mass_Proporties, using ACIS7.0 calculates center of gravity GP of aggregate particle physical model MM,
M is moved to GPMThe position overlapped with global coordinate system initial point, making rotary shaft is x-axis；
2) the bottom modeling function Get_Bounding_Box, using ACIS7.0 obtains the current 3D bounding box 3DBeBox of MMAnd
Calculate its volume BeVMBoxAfter, with GPMCentered by, make M rotate forward θ=1 ° around rotary shaft, it is thus achieved that M postrotational 3D bounding box
3DAfBoxMAnd calculate its volume AfVMBoxIf, BeVMBox>AfVMBox, by the orientation before M rollback to rotation, and set rotation side
To for rotary shaft forward, otherwise make M around rotary shaft counter-rotating θ=1 °, if now BeVMBox>AfVMBox, by M rollback to rotating
Direction of rotation is set reverse as rotary shaft behind front orientation, if now BeVMBox<AfVMBox, rotation terminates, and performs step 5)；
3), with GPMCentered by, along step 2) direction of rotation that sets, after determining actual θ value according to 3D corner decay rule, by M
Rotate θ angle；
4), judge whether 3D terminating rotation condition is reached, if so, rotate and terminate, perform step 5), otherwise perform step 3)；
5) making rotary shaft, successively is y-axis and z-axis, performs step 2)～4), terminate until M all rotates around three coordinate axess, it is thus achieved that
The 3D minimum bounding box 3DMinBox of MM；
Wherein:
That described 3D bounding box refers to be surrounded aggregate particle physical model M completely, each limit is all parallel with a certain coordinate axes
Minimum cuboid；
Described 3D corner decay rule refers to differentiate that M surrounds at the 3D not rotated, rotate under θ angle and three kinds of situations of θ+1 angle
Box body is long-pending to successively decrease the most successively, if so, determines that the anglec of rotation is θ, if it is not, again differentiate after then being halved by θ, until obtaining certainly
Differentiation result or θ decay to 1 °, wherein the initial value of θ is set to 8 °；
Described 3D terminating rotation condition refer to M not rotating, in the 3D bounding box volume that rotates under 1 ° of two kinds of situation, the latter is relatively
Greatly；
Described 3D minimum bounding box 3DMinBoxMRefer to that M carries out any direction with any center around any axle, revolves at any angle
Volume reckling in obtained 3D bounding box after turning.
Aggregate size three-dimensional method for automatic measurement based on CT image the most according to claim 1, it is characterised in that: described
In step (4), it is described as follows step based on the aggregate size measuring method controlling plane:
1) the 3D minimum bounding box 3DMinBox of aggregate particle physical model M, is obtainedMIn longest edge one, it is judged that and determine with
Parallel coordinate axes AxisL；
2), use ACIS7.0 bottom modeling function Section, with 3DMinBoxMAxis minimum in 8 summitsLCoordinate figure
For AxisLOn original position, maximum AxisLCoordinate figure is AxisLOn final position, along and AxisLVertical direction,
To be spaced intervalMM is carried out cutting, it is thus achieved that q part Seck(k=1,2 ..., q), and use the bottom of ACIS7.0 to build
Mould function Mass_Proporties calculates each SeckCenter of gravity
3), calculating is derived fromExtremelyVectorial Veck,k+1(k=1,2 ..., q-1), use the bottom of ACIS7.0 to build
Mould function Section, respectively edge and vector Vec(k,k+1)(k=1,2 ..., q-1) vertical direction, mistakeIntercept the cross section of M, obtain the set SFS={SF of FACE style cross-sectionsk| k=1,2 ..., 2q-2}；
4) the bottom modeling function Get_Face_Normal, using ACIS7.0 obtains SFS middle section SFk(k=1,2 ..., 2q-
2) normal vector, it is judged that itself and coordinate axes AxisLAngle between forward, obtains SFkA summit and centered by it, rotate
SFkTo its normal vector and AxisLParallel orientation；
5) the bottom modeling function Get_Face_Box, using ACIS7.0 obtains SFkCurrent 2D bounding boxMeter
Calculate and obtain its centerAnd areaAfter, by SFkMove toThe position overlapped with global coordinate system initial point, withCentered by, make SFkAround AxisLRotate forward α=1 °, it is thus achieved that SFkPostrotational 2D bounding boxAnd calculate
Its areaIfBy SFkOrientation before rollback extremely rotation, and set direction of rotation as AxisL
Forward, otherwise by SFkSF is made behind orientation before rollback extremely rotationkAround AxisLCounter-rotating α=1 °, if nowThen by SFkDirection of rotation is set as Axis behind orientation before rollback extremely rotationLInversely, if nowRotation terminates, by SFkCurrent 2D bounding box as its 2D minimum bounding boxPerform
Step 8)；
6), withCentered by, along step 5) direction of rotation that sets, after determining actual α value according to 2D corner decay rule,
By SFkRotation alpha angle；
7), judge whether 2D terminating rotation condition is reached, if so, by SFkCurrent 2D bounding box as its 2D minimum bounding boxOtherwise perform step 6)；
8), to each cross section SF in SFSkPerform step 5 successively)～7), until obtaining setThe relatively longer sides length of side of each element in set, by unit the longest for the longer sides length of side
Cross section corresponding to element is as controlling plane CF, and the corresponding longer sides length of side is sieved size as the minimum gathered materials, with closest
And the standard screen screen size being not less than minimum size of sieving is aggregate size；
Wherein:
Described cross section SFkNormal vector refer to and cross section SFkVertical vector；
Described 2D bounding box refer to limit that the plane that parallel with a certain coordinate surface is surrounded completely, each all and certain
The minimum rectangle that one coordinate axes is parallel；
Described 2D corner decay rule refers to differentiate SFk2D not rotating, under rotation alpha angle and three kinds of situations of α+1 angle surrounds
Box area successively decreases the most successively, if so, determines that the anglec of rotation is α, if it is not, again differentiate after then being halved by α, until obtaining certainly
Differentiation result or α decay to 1 °, wherein the initial value of α is set to 8 °；
Described 2D terminating rotation condition refers to SFkNot rotating, in the 2D bounding box area that rotates under 1 ° of two kinds of situation, the latter is relatively
Greatly；
Described 2D minimum bounding boxRefer to SFkThis center and SF is walked around with any centerkVertical rotary shaft
Carry out any direction, rotate at any angle after area reckling in obtained 2D bounding box；
Described control plane CF refers to determine to gather materials particle entities model M can be by the cross section of minimum sieve aperture；
Described minimum size of sieving refers to the minimum sieve hole dimension that can pass through that gathers materials；
Described aggregate size refers to gathering materials the minimum sieve hole dimension that can pass through in standard screen；
Described standard screen refers to be made up of the bushing screen in one group of multiple different size aperture, for sieving the work that different size gathers materials
Tool, screen size is followed successively by from small to large: 1.18mm, 2.36mm, 4.75mm, 9.5mm, 12.5mm, 19mm.
Aggregate size three-dimensional method for automatic measurement based on CT image the most according to claim 1, it is characterised in that: described
ACIS7.0 is the THREE DIMENSION GEOMETRIC MODELING engine based on Object Oriented technology produced by Spatial company of the U.S..
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610195893.9A CN105954161A (en) | 2016-03-30 | 2016-03-30 | CT-image-based three-dimensional automatic measurement method for particle size of aggregate |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610195893.9A CN105954161A (en) | 2016-03-30 | 2016-03-30 | CT-image-based three-dimensional automatic measurement method for particle size of aggregate |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105954161A true CN105954161A (en) | 2016-09-21 |
Family
ID=56918054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610195893.9A Pending CN105954161A (en) | 2016-03-30 | 2016-03-30 | CT-image-based three-dimensional automatic measurement method for particle size of aggregate |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105954161A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108489872A (en) * | 2018-03-23 | 2018-09-04 | 奥星制药设备（石家庄）有限公司 | Online granularity monitoring method and system |
| CN109191571A (en) * | 2018-09-30 | 2019-01-11 | 华南理工大学 | A method of gather materials using 3D printing technique preparation mechanical test standard |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101403683A (en) * | 2008-11-17 | 2009-04-08 | 长安大学 | Method for analyzing porous asphalt mixture gap structure by using CT technology |
| CN102760309A (en) * | 2012-05-30 | 2012-10-31 | 合肥工业大学 | Bituminous pavement test piece X-ray CT (computed tomography) image-based aggregate mesoscopic solid model reconstruction method |
| CN104599321A (en) * | 2015-01-24 | 2015-05-06 | 合肥工业大学 | Building method of discrete element module of true aggregate particles based on X-ray CT image |
-
2016
- 2016-03-30 CN CN201610195893.9A patent/CN105954161A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101403683A (en) * | 2008-11-17 | 2009-04-08 | 长安大学 | Method for analyzing porous asphalt mixture gap structure by using CT technology |
| CN102760309A (en) * | 2012-05-30 | 2012-10-31 | 合肥工业大学 | Bituminous pavement test piece X-ray CT (computed tomography) image-based aggregate mesoscopic solid model reconstruction method |
| CN104599321A (en) * | 2015-01-24 | 2015-05-06 | 合肥工业大学 | Building method of discrete element module of true aggregate particles based on X-ray CT image |
Non-Patent Citations (2)
| Title |
|---|
| 孔垂品 等: "模具零件的最小包围盒生成算法", 《模具工业》 * |
| 段跃华 等: "基于CT断层扫描图像的混凝土粗集料三维虚拟筛分", 《吉林大学学报（工学版）》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108489872A (en) * | 2018-03-23 | 2018-09-04 | 奥星制药设备（石家庄）有限公司 | Online granularity monitoring method and system |
| CN108489872B (en) * | 2018-03-23 | 2022-03-04 | 奥星制药设备（石家庄）有限公司 | Online granularity monitoring method and system |
| CN109191571A (en) * | 2018-09-30 | 2019-01-11 | 华南理工大学 | A method of gather materials using 3D printing technique preparation mechanical test standard |
| CN109191571B (en) * | 2018-09-30 | 2023-09-12 | 华南理工大学 | Method for preparing mechanical test standard aggregate by applying 3D printing technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107679441B (en) | Method for extracting height of urban building based on multi-temporal remote sensing image shadow | |
| CN106600690B (en) | Complex building three-dimensional modeling method based on point cloud data | |
| CN105045950B (en) | A kind of bridge security assessment system based on 3 D laser scanning | |
| Chen et al. | Automatic extraction of discontinuity orientation from rock mass surface 3D point cloud | |
| CN106327573B (en) | A kind of outdoor scene three-dimensional modeling method for urban architecture | |
| CN109949326B (en) | Building contour line extraction method based on knapsack type three-dimensional laser point cloud data | |
| CN112525164B (en) | Method for detecting deformation of super high-rise building based on unmanned aerial vehicle oblique photography technology | |
| CN109492852A (en) | A kind of detection method for quality of the water conservancy project structure based on BIM | |
| CN106091972B (en) | A kind of building change detecting method projecting dot density based on moving window | |
| CN104809756B (en) | Asphalt mixture gap spatial construction method based on X-ray CT images | |
| Jin et al. | Aggregate shape characterization using virtual measurement of three-dimensional solid models constructed from X-ray CT images of aggregates | |
| CN103714571A (en) | Single camera three-dimensional reconstruction method based on photogrammetry | |
| CN105180890A (en) | Rock structural surface occurrence measuring method integrated with laser-point cloud and digital imaging | |
| CN104599321A (en) | Building method of discrete element module of true aggregate particles based on X-ray CT image | |
| JP2013072704A (en) | Extraction method of discontinuity surface of tunnel working face and device of the same | |
| CN108846860B (en) | A kind of damaged cylindrical drainage pipeline inner wall three-dimensional rebuilding method | |
| CN104036538A (en) | Method and system for reconstruction and analysis of earth-rock aggregate three-dimensional microstructure | |
| CN104331922A (en) | Three-dimensional reconstruction method for soil microstructure | |
| Chmelina et al. | A 3-d laser scanning system and scan data processing method for the monitoring of tunnel deformations | |
| CN105550428A (en) | Bridge security evaluation method based on TLS (three-dimensional laser scanning) technique | |
| CN105205855B (en) | A kind of soil erosion situation method for measuring | |
| CN103017734A (en) | Pole and tower gradient of slope measuring method based on laser radar | |
| CN102944566A (en) | Method for testing digital imaging accuracy of bituminous mixtures based on computed tomography (CT) machine | |
| CN105954161A (en) | CT-image-based three-dimensional automatic measurement method for particle size of aggregate | |
| Liu et al. | Evaluating angularity of coarse aggregates using the virtual cutting method based on 3D point cloud images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Jin Can Inventor after: Wang Peisong Inventor after: Zhang Weihua Inventor after: Liu Kai Inventor after: Wang Xinlei Inventor before: Jin Can Inventor before: Zhang Weihua Inventor before: Wang Peisong Inventor before: Liu Kai Inventor before: Wang Xinlei |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160921 |