A Practical Method of
Remotely Sensed Digital Image's Resolving Power
Izumi Kamiya
Geographical Survey Institute of Japan
Kitasato-1, Tsukuba-shi, Ibaraki-ken
Japan
Commission I
1. Abstraet
A method which decides resolving power of digital remote
sensing sensor in field experiment was developed. It was
applied to MOS-1 airborne verification experiment and predicted
a resolving power of MESSR on MOS-1 satellite. :
An outline of this method is followins. An airphoto sjienarl,
refrectancs of which is like step. function, is.laid slightly
obliquely (angle &) to perpendicular to flying direction and
observed (Fig. 1). It is equivalentobouwobsenving«a "object,
radiance of which changes like step functions: in’ various
phases. A calculation method of PSF and MTF from the data was
developed.
2. MOS-1 and MESSR
NASDA(National Space Development Agency of Japan) launched MOS-
1(Marine Observation Satellite. LY dan 92 Le BELO Sr LIS
launching, NASDA conducted MOS-1 airborne verification experi-
ment in order to test sensor equipment. GSI(Geographical Survey
Institute of Japan) took part in the experiment, and evaluated
resolving power of MESSR(Multi-spectral Electric Self-scanning
Radiometer). It is & visible and infrared image sensor on the
MOS-1. Table 1 is selected technical data of the MOS-1 and the
MESSR.
The MESSR employed linear CCD(Charge Coupled Device) sensor
like SPOT HRV. One of the most significant difference between
linear CCD sensor and seanner(ex. LANDSAT MSS, TM) is follow-
ing. Seanner samples incident light in IFOV(Instantaneous Field
of View) in a moment. But CCD accumuliatesiincddent:lightitin
IFOV while satellite moves corresponding length between two
consecutive lines. So, the area that a detector observes
changes during the accumulation. It contributes: to lower
resolving power in flying direction, but no ‘contribution in
perpendicular to the flying direction.
3. General idea of Ihe method
In order to obtain PSF(Point Spread Funetion) in certain direc-
tion, it is enough to observe airphoto.signal, refrectance..of
which changes like step function in the direction, in various
locations of sensor relative to the edge. But it is impossible
to design a airplane or satellite based experiment to fix a
pixel to certain location. Also it' is very diffieult to observe
the Locations of pixeles in sub-pixel order. So, statistical
processing of edge observation data, in which every locations
were thought to appear in same probability, was used for field
64