2-2-2 MOVEMENT CHANGES AROUND
THE PITCH AXIS
This appear as a change in the direction of
the y-axis that is perpendicular to the scan
direction of CCD. Thus, the equation 1s:
kw, = Ho. (5)
2-2-3 MOVEMENT CHANGES AROUND
THE YAW AXIS
Movement changes around the yaw axis
rotates the imaging position; as shown in
Fig.2(c). Since this is very small angle change it
can be represented by the following equation if
L isset as an observation width on earth:
kw: =(L/2-x)x(1-cosd)
=(L/2-x)& /2, (6)
EM; = (1 /2- x)ó.- CT
In the high resolution earth observation
system, the normal H number is 800 km (832
km for SPOT satellite ); L is smaller than
200 km (60 km for SPOT satellite): the effective
changes around the yaw axis is relatively small
compared to the movement changes of the roll
and pitch axis.[3]
2-3. EFFECT OF FLUCTUATION FOR TDI
The image pick-up position of the CCD is not
completely same because of satellite's attitude
fluctuation.
Considering the interference from contiguous
picture elements, 2-D repeating high-frequency
input signals are most sensitive to the attitude
fluctuation. Therefore, we use 2-D sinusoidal
wave pattern defined by
|l * cos(2zf,x)cos(2z/,y)
3
fx,y)Yys (8)
where f, and f, are the repeating signal
frequency in the x direction and in the
y direction. Fig.3 shows the simulation pattern
generated by Eq.(8).
In this case, the amplitude component of a sin
signal at 2xf:= 2mz and 22f: = (2m+1)æ
( m; integer ) is represented by
Sue rS Hy umor E cos(2ik.)
Msn zn Poesia p: ©)
Tc:Ww,'3p,
where
ki fkuw xt fekywW y. (10)
k = fxkxw x Fr fykyw JA. (11)
Ciz fujps- Kw (12)
és: = Jfp,tkiií. (13)
| —$— without fluctuation |
| —3— with fluctuation |
| MO
O7 Dr
= —
» 6 FE
+
S 5
1
| 3 3
| =
E 2
> 1
N
(008
| 0 10 20 30 40
| times
Fig.4 The result of performing TDI with Eq.(8).
The simulation result of performing TDI with
Eq.(8). From Fig.4, we can't acquire the S/N
improvement ratio more than 3 without
detection and correction of the satellite attitude
fluctuation.
18 Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998
3-1.
mov
ante
poin
dete
errc
f
we (
I
Fro
atti
Anter
CCD
Fig
3-2
arc
thi
wh
nu
sm
fir
sh
ou