These planned systems may be compared with those of LANDSAT-1 and 
-2 using the EIFOV concept. Referring to Fig. 11, MTF's have been cal- 
culated for detectors with IFOV's of 238, 120, 79 and 30 m respectively 
and the corresponding spatial frequencies for the 50 percent modulation 
point noted as 8, 17, 26 and 66 cy/mm for an image scale of 1:3,369,000 
(70 mm format). If visual interpretability is assumed to increase lo- 
garithmically with linear increases in image resolution (as with photo- 
graphic systems), an approximate 40 percent gain in image interpretability 
will be realized by improving the IFOV by factors of 2 to 3. Although 
these calculations are approximate, they do provide a basis for estimating 
the potential applications of the LANDSAT data. 
A similar approach can be utilized to compare the various LANDSAT 
sensors with the photographic systems employed on Skylab. In Fig. 12, 
for example, MTF's for the LANDSAT and high-resolution black-and-white 
Skylab sensor systems have been adjusted to the 1:1,000,000 image scale 
planned for future first-generation LANDSAT products (NASA, 1976). Based 
on these MTF's, the images provided by the 30 m IFOV of the Thematic 
Mapper will be of slightly higher quality than equivalent scale second- 
generation products from the S-190A MPF. 
MTF's have also been employed by Slater (1975b) to compare two 
hypothetical earth satellite systems with equivalent EIFOV's: a photo- 
graphic system providing analog imagery and an electro-optical system 
in which a solid state array of CCD's replaces the film and provides 
digital output. This study demonstrates the ground resolution advantages 
of the photographic system and its superiority for cartographic tasks re- 
quiring the visual examination and mensuration of imaged objects. The 
electro-optical system, however, has a better signal-to-noise ratio re- 
sulting in improved discrimination of small spectral reflectance differ- 
ences. Consequently the electro-optical system is recommended for remote 
sensing applications in the earth sciences, whereas the photographic system 
is preferred for cartographic tasks. 
CONCLUSION 
In the four years since the XII Congress, the utilization of OTF/MTF 
in the design and evaluation of photographic and electro-optical remote 
sensor systems has increased significantly despite the requirements for 
complex instrumentation and analysis procedures. For photogrammetric tasks, 
however, these (and other) measures of system performance must be trans- 
lated into measures of quality which relate to the interpretability and 
measurability of image detail. Today, this problem is magnified by the 
fact that photogrammetrists are receiving image data recorded in analog 
and digital formats by diverse photographic and electro-optical systems. 
Scales vary considerably, as do the measures of performance and quality. 
The term "resolution," for example, may refer to lpr/mm, TV lines, IFOV, 
EIFOV, meters/bar, meters/lpr, etc., depending on the system being dis- 
cussed and the reference background of the discussant. Consequently, it 
appears appropriate to recommend that the OTF/MTF Working Group be expanded 
to an Image Quality Working Group. Investigations of measures of system 
performance and image quality, and their relationships to photogrammetric 
problems could be the assigned tasks of this proposed working group.