Full text: XVIIIth Congress (Part B2)

  
SEMI-AUTOMATIC DIGITAL PHOTOGRAMMETRIC SYSTEM ON PC 
OYAMA Yoichi 
Kokusai Kogyo Co., 
KEYWORDS: 
ABSTRACT 
Ltd., JAPAN 
DTM, Orthoimage, Stereo Matching, Correlation, Personal Computer. 
Many studies have been made so far on the automatization of photogrammetric system through dig- 
ital processing. Some systems based on workstation have been developed in recentyears and are 
now entering the stages of practical use, but there are still many problems left unsolved such 
as the difficulty of complete automatization and complexity of processing procedures. 
In this study under such circumstances, we aimed at a simple system based on personal computer 
and developed the Personal Digital Photogrammetric System "PyramidStereon' employing "semi- 
automatic" processing which jointly uses human eyes and hands at the areas where the automati- 
zation is difficult. This paper will report the process of system development and verification 
results of test examples, and discusses the practical utility of the system. 
1. Outline of past studies 
The studies have already been made since 1960's 
on the automatization of photogrammetric system, 
. especially on the compilation of DTM from 
stereo aerial photographs. In these studies, 
the automatic search (stereo matching) of cor- 
responding points has been made mainly by 
"image correlation method” which compares the 
similarity of left and right images. 
From 1960's to 1980's, the studies were made 
for the development of dedicated hardware call- 
ed the Auto Correlator. The representative 
system of this development would be the Gestalt 
Photomapper (Hobrough, G.L., 1970). This was a 
system to identify the correlation among the 
photographic data of left and right photographs 
on the electric circuits, and was used as the 
basic method of image processing of later days. 
In and after 1980 s, the experimental systems 
were successively developed, which verify the 
processing result of image correlation method 
by workstations. The methods such as data input 
method, image display, rectification and stereo 
matching have been established through these 
development processes. Especially, the method 
to implement one dimensional search on epipolar 
line after the rectification was said to be ef- 
fective for stereo matching, and multi-step 
searching methods using the image pyramid were 
employed to many system to shorten the proces- 
sing time and prevent the mis-correlation. On 
the other hand, however, many problems were 
pointed out relating thereto such that there 
were difficulty in detection and matching of 
occulsion portions or deviation was observed 
comparing with existing mapping method because 
only the tree-top altitudes were obtained by 
matching. These problems were coped with and 
efforts were made to develop the system of 
practical utility by introducing the notion of 
so-called "man-machine interface," namely to 
use the help of human eyes and hands in stead 
of perfect automatization. There is EW: General 
Dynamics Experimental Workstation (General Dy- 
namics, 1985) as a representative example of 
this type. 
As the ISPRS Kyoto Conference held in July, 
1988 as aturning point, many workstations were 
manufactured aiming at the practical use. The 
representative systems were DSPl: Digital 
Stereo Photogrammetric System (Kern Inc., 1988), 
HAI-750 (General Dynamics/Helava Assoc. Inc., 
1990), DPW750 (Helava Assoc. /Leica Inc. , 1992), 
PHODIS (Carl Zeiss Inc., 1992) and PhotoScan 
(Intergraph Inc. 1992). 
As stated in the above, the studies on the 
automatization of photogrammetric system have 
been made mainly around the image correlation 
method, and many kinds of systems, mainly the 
288 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B2. Vienna 1996 
  
WOI 
mer 
rel 
met 
dur 
it 
acc 
sib 
by 
ram 
ope 
lik 
As 
ima 
cas 
rel 
pho 
pro 
ste 
and 
ple 
sys 
pri 
wor 
tha 
Bas 
sta 
Pho 
is” 
whi 
of | 
PUT] 
pile 
sem 
dif: 
Ware 
fig
	        
Waiting...

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.