cedure
r orientation
planning and
pilot control
GmbH. High
during flight,
are obtained
ocessing, the
solution still
video image
oint tracking
acking result
nate on high
matching of
; of this sub-
ass points. In
r orientation
The result of
of candidate
jn and scale.
e positioning
itrol points is
st regression.
ed based on
delling from
lon
tation
ure point and
Code” (OC)
f OC image
nstead of the
ization of the
st pixel. The
a ed
tan AR
y ;
Al |+|A | > (1)
se if |ar|+|ar,|27
Cy =
2x ;
=== otherwise
AO
In this equation, 4/, and AZ, show horizontal and vertical
gradient of pixel (x, y). N is the quantization level of direction.
N is set to 16 typically (Figure 3). The “y” is the threshold value
for the suppressing of small gradient pixel.
%
iue
s Ford i 3x
Figure 3. Orientation Code (N = 16)
In order to extract common feature points from OC image, we
use Orientation Code Richness (OCR) that extracts the pixels
which have high entropy of OC. The entropy of the local
limited area of the pixel size M-by-M region at the interest pixel
(x, y) is calculated as follows,
P.) die /M*—h,(N) i
E, =) P,(i)log, P,(i)
i=0
Where h,,(i) (i =0,1,,,,,,N-1) means frequency of OC of M-by-M
pixel size region.
When each OC goes with uniform distribution P,,(i)- 1/N, the
maximum value of entropy £,,. is log;N. Consequently, the
richness R,, is defined as
E um if E,2«,E 3
pay o ob m IS E
R,, = E x = OE, x y e" max ( )
0 otherwise
Where, the threshold value a, is defined to remove low entropy
area.
The matching process using OC image (OCM) is similar to
other simple image based template matching. The difference
between a template image patch from OC image (Ot) and search
OC image (Oi) is defined as following equation,
=-L 40,0)
M M
Zo) min{a—&|, N —|a— 4} if a#N,b#zN (4)
N/4 otherwise
where D - difference between O, and Oj
M = Size of template image patch
d = difference between a pixel and b pixel
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B7, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
Figure 4 shows the example of OC image. From this image, it is
understood that the OC image is independent from the changing
brightness and also OC image describe the important feature of
image. Figure 5 shows the result of OCR. The bright pixel on
OCR means the high entropy of OC. Figure 6 shows the
example of OCM. From the robustness of OCM, almost points
are not influenced by the changing of brightness or noise of
image.
8 E
a)Original Image (8bit Gray Scale)
segs
b) OC Image
Figure 4. Example of OC image
Figure 5. Example of OCR
