t of non-contact
ured object for
ce, the projector
calibration of a
The planar grid
:d or its intrinsic
] to calibrate the
'€ coordinates of
the intrinsic and
ic and extrinsic
n parameters are
t forward in this
The camera has
ters are known
ound to provide
rrespondence of
isformation, the
ked out. By this
| the. expensive
posed technique
ted by the space
sing in future of
> feasibility and
inique proposed
ita.
It can be used to
ts as the same
ly the points of
are lack of the
of the measured
insic parameters
the calculation
UTHM
inar grid are the
the projector, a
target grid slide is designed first so that the coordinates of its
grid points can be acquired as the known data. The camera
requires to have been calibrated already or its intrinsic
parameters are known in advance. The planar grid is also
designed first to be able to obtain the coordinates of the points
of the grid.
First, the positions of the projector and the camera are adjusted
to two appropriate places. The camera is focused to take clear
photos of the planar grid and the projector is also focused to
project clearly a target grid slide onto the plane. Second, the
camera is used to take an image of the planar grid. Then the
planar grid is covered by a white paper (or other things), which
makes the planar grid seem to be a plane. The projector
illuminates the target grid slide onto the plane. The camera is
used again to take the images of the grid illuminated. In this
process of taking the image of the illuminated grid, it is
important that the positions of the camera and the planar grid
are both never changed. Using the image processing method,
the coordinates of the grid points of the image of planar grid are
extracted. Because the space coordinates of the planar grid
points are known as design, the extrinsic parameters of the
camera can be calculated by the space resection method. So that
the intrinsic and extrinsic parameters of the camera have been
gotten at present. Using the image processing method again, the
coordinates of the points of the image of the projected grid are
extracted, too. By the collinear equations, the space coordinates
of the projected points of the image are calculated in reverse,
because the Z value of the space coordinates of these points is
zero always. For the projector, the space coordinates of the grid
points projected have been acquired by the process above, and
the image coordinates of these points are known as design in
advance. Using the correspondence of 2D-DLT and collinear
equation, the decomposition of initial values of the projector
intrinsic and extrinsic parameters is deduced. Third, the
positions and orientations of the projector and the camera are
changed to take another two images as the second step. At least
all images need to be taken from two different positions and
orientations. Then at least two sets of the initial values of the
projector parameters can be computed out. Finally, the intrinsic
parameters of the projector can be worked out by the whole
adjustment based on these initial values above. So the ordinary
projector has been calibrated entirely by this way.
2.2 Algorithm
The collinear equations are:
X—x =-f a (X X, )+6,(Y-Y)+ce, (Z-Z,)
s a x )50.(Y 27 Yee lZ-7) 2i
aX - X, )* b,(Y - Y.)«c,(Z - Z.)
aX - X,)«5,(Y - Y,)« e, (Z - Z,)
y-y.m-f
where f, xo, yg — the intrinsic parameters of the projector
Xs» Ys, Zs = the coordinates of the projector centre
X, Y, Z = the space coordinates of points
x, y = the image coordinates of the relative points
R (aj bj c; i71,2,3) 7 the rotated matrix made up of
rotated angles 9, Q,K
Z=0
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004
2D-DLT is expressed as:
ho X c h.Y oh,
ho Xon Y]
EX UY C.
Abel
x =
where H;(i=1,2...8) = parameter
X, Y = the space coordinates of the points projected
x, y = the image coordinates of the relative points
When the number of points observed in an image is more than 4
the parameter H; can be calculated out.
3
According to the formula (2), formula (1) can be transferred
into formula (3):
a d Ds b ir I T
" (/ a -Sx je i Ba) s - aX, LA )
= 3
Lx ya 3)
A A
Math Th Led +6,
( Bey ore Ha o ax, +b sez)
“Bx fy
A
y =
Compared formula (2) and (3), then:
a, d,
n ER
(4)
ho
hf
AA
h, - fy
Aoi Ao (5)
| +0, by 4
fs = [=k
$
h =x, Ar +bY +¢,Z.)
^ (6)
fa = Vo za, t b,Y, - c,Z,)
a,
hkE-—
; n
h,--—