2004 
MULTI-TRIANGULATION TO GET GCP 
FOR OLD UNPREMARKED AERIAL PHOTOGRAPHS 
Fahmi Amhar? 
* Researcher in Mapping & Technical Manager of Kalimantan Mapping 
National Coordinating Agency for Surveys and Mapping, Bakosurtanal 
Cibinong Science Centre, JI. Jakarta Bogor Km. 46 Cibinong, Indonesia 
Telp. /Fax. +62-21-87901254; email: famhar@telkom.net, famhar@yahoo.com 
TS 1 YF Youth Forum - Photogrammetry 
KEY WORDS: GPS/INS, georeferencing, identification, triangulation, photographs, orientation, adjustment 
ABSTRACT: 
If a large number of old aerial photographs is exist but there is no premarked ground-control- and also no kinematics-GPS on board 
of photo-centre-position for aerial triangulation, then the post GCP-survey would be very hard. The most problem is to identify the 
place on photographs to the current situation and to get the place. In the first part, the experience of such project in Kalimantan 
will be explained. In the second part is the idea how to accelerate and optimised the whole GCP survey. 
1. BACKGROUND 
1.1 General Condition 
Kalimantan (Borneo) is one of the largest islands in the world. 
The Indonesian part of Kalimantan is about 540,148 sqkm. 
The Island is just 43 % topographic mapped on 1:50.000 scale. 
A large part is not yet mapped due to: (1) the ever-cloudy 
situation; (2) the limited state budget to do systematic mapping 
over the whole island. 
During the Era of Soeharto's government, the ministry for 
forestry has obligated to all forest concession holder company 
to make aerial photographs for their concession area. Between 
1990-1993, each company made their photographs used 
different camera, aircrafts etc. These aerial photographs are 
stored to the ministry as evidence of the forest situation. No 
topographic map is made from these photographs. 
After the end. of Soeharto's regime, there is new condition, 
which focused to re-use the photographs. The paradigm of 
district wide-autonomy has increased the consciousness of the 
district authorities of Kalimantan, that they need also qualified 
topographic map in middle and larger scale. This demand is 
pushed by the facts of increasing illegal logging, wild forest 
fire, and claim from indigenous people for land-common-right 
and other environmental problem. These entire problems are 
acute. 
There are some technology on the market to make 
topographical map for scale 1:50.000, ie. (1) airborne 
photogrammetry using existing (forestry) photos; (2) airborne 
photogrammetry using new photos; (3) SPOTS-Stereo or 
ASTER-Stereo; (4) Airborne Interferometric SAR (IFSAR); 
and (5) Lidar Imaging. Due to the budget constraints, and the 
cloud situation, the option (1) is in favor. 
The reason: topography data (DTM) is the most expensive part 
in topographic database and in the area is mostly not changed. 
Only the land cover is changed, and this should be updated 
from Landsat or Aster data. The option (1) shows the respect : 
of the current government to save money and re-use the 
existing assets. The problem is this concept never proven for a 
large area. ; 
In 2002 the National Coordinating Agency for Surveying and 
Mapping (Bakosurtanal) decided to use aerial 
photographs to make new topographic map for part of East 
Kalimantan, about 80,000 sqkm (8 Million hectare) to execute 
in 2003. The financing plans about 14000 aerial photographs 
to be processed in this year. : 
1.2 The Selection of the Photographs 
About 70000 aerial photographs of scale 1:15.000 — 1:22.000 
were made for forest concession holder. The large scale is 
while the aircraft had to fly under the cloud. The photographs 
would be made mainly on forest area; also not what in the time 
of capture was cities, swamps, big river etc. Different holder 
means different project, different camera, different scale, 
different numbering system on photos and different photo 
quality. It was very hard job to select adequate and useful 
photos. There are many overlaps between photos. In other 
location, there are also individual gaps where no any photo 
exists. The overlaps are caused by the fact that some company 
has taken photographs several times, by some reason 
(unsatisfying of the previous result, regular update to know the 
last forest situation after fires, cut off or replant). The last and 
best quality photo will be taken to the next process. 
The selection process used the photo-index database. 
Unfortunately, this database was in many cases inaccurate.