004 
ong 
the 
INE 
the 
s In 
x the 
eans 
than 
1 the 
ore, 
the 
the 
cond 
hood 
ingle 
rage. 
f the 
ML 
jifler, 
fier. 
ed as 
ained 
e two 
sifier 
ig Set 
als in 
1 the 
of the 
Upon 
ssifier 
cctral 
sifier, 
d and 
yectral 
mage. 
ogged 
lied to 
es and 
ion of 
ctance 
ich as 
n used 
based 
orward 
on the 
le with 
we that 
ifier 18 
nce in 
an get 
ing the 
image 
is take 
of the 
ximum 
ise and 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
use of this new software. More importantly, the increasing 
number of illegal single tree felling RKL1 makes it necessary to 
have to an accurate method for detection of this type of logging 
  
NLP Detection 
(SP Classifier versus ML Classifier) 
    
  
   
  
     
       
Legend 
SP detection of NLP 
[J ML detection of NLP 
Wl Common detection of NLP 
E] Common detection of Other 
  
4 km 
  
  
  
Figure 13. Comparison of NLP Detection by SP versus ML 
Classifier. 
TET 
ri 
E gu €. . a 
ESI = 
= = 
   
  
Legend 
SP detection of NLP 
[J ML detection of NLP 
Wl Common detection of NLP 
EZ] Common detection of Other 
  
Figure 14. Subset of Map showing NLP detections by ML and 
SP Classifier. 
S. CONCLUSIONS 
The results of this study showed that single tree felling can be 
detected using Landsat-7 ETM+ image and the IMAGINE 
Subpixel Classifier. Detection was studied using logged points 
that were less than a year old. In addition, the use of GIS and 
other ancillary data combined with expert knowledge can help 
improve the result of image classification as well as 
characterize the felling as planned or illegal. The findings are 
listed according to the research questions stated in chapter one. 
The IMAGINE Subpixel classifier produced a higher accuracy 
compared to the Maximum Likelihood Classifier in detecting 
single tree felling in the tropical forest using Landsat-7 ETM+ 
image. 
939 
6. REFERENCES 
BFMP. (2002, May 17, 2002). Labanan Concession Re- 
allocation. Retrieved July 12, 2003, from the World Wide Web: 
http://www .bfimp.or.id/Publications/Labanan_brief notes03.htm 
Bhandari, S. P. (2003). Remote Sensing for Forest Certification: 
Detecting and Characterizing Selective Logging in Tropical 
Forest: a case study in Labanan concession, East Kalimantan, 
Indonesia. Unpublished MSc., ITC, Enschede. 
Sist, P., ef al. (2003). Reduced-impact logging in Indonesian 
Borneo: some results confirming the need for new silvicultural 
prescriptions. Forest Ecology and Management, 179(1-3), 415- 
427.