o these 
ion is 
re 4). 
and 
ape of 
tilized 
curacy. 
gorithm 
with 
each 
to the 
ls 
| of 
25.0. 
ined by 
| pixel 
celihood 
on, 
ation 
cedure, 
uni form 
Or 
texture 
of a 
n Figure 
ssifier. 
e, 
t change 
E cdi 
  
regularity of change patterns(Table 2). With the aid of this 
classifier, processing times largely decreased and classification 
accuracy has increased. 
V. RESULTS 
The classified results is shown in Figure 6. A small 
misclassifications, e.g. noises near the boundaries were manually 
corrected pixel-by-pixel basis. Since three land use categories 
(sea water, seashore, plantations) do not appeare in this target 
area, the number of classification categories were twelve. 
Table 3 represents percentages and change ratios of each land 
use categories between 1971 and 1979. 
VI. CONCLUSIONS 
AS a result of this study, the following conclusions were 
obtained. 
1) The renewal system of NNLI, which utilizes remote sensing 
data(LANDSAT/ airborn data) and digital processing techniques, 
was established. The presented renewal method is more 
economical than conventional manual interpretation method. 
2) Conventional maximum likelihood classification algorithm can 
not acheve sufficient classification accuracy for land use(not 
land cover) classification. 
3) Histgram matching algorithm is highly useful for land use 
classification, although it consumes more times than maximum 
likelihood algorithm. 
VII. REFERENCES 
l. L.D.Miller,K.Nualchawee and C.Tom, "Analysis of the Dynamics of 
Shifting Cultivation in the Tropical Forests of Northern 
THailand Using Landscape Modeling and Classification of LANDSAT 
Imagery" ,Proc. 12th Inter. Sympo. on Remote Sensing of 
Environment (ERIM), p.1167-1185 (1981) 
2. D.M.Carneggie,C.S.Holm, "Remote Sensing Techniques for 
Monitoring Impacts of Phosphate Mining in Southeastern Idaho" 
Proc. 2nd Annual Pecora Sympo. (1976) 
3. K.Fukue,H.Shimoda and T.Sakata, "Radiometric Correction for 
Aerial Photography", Journal of the Society of Photographic 
Science and Technology of Japan,Vol.43,No.6 (1980) 
4. S.Hsu,"Texture-Tone Analysis for Automated Land-Use Mapping", 
Photographic Engineering and Remote Sensing, Vol.44 
p.1393-1404 (1978) 
5. J.C.Coiner,R.C.Bruce,"Use of Change Detection in Assessing 
Development Plans : a Philippine Example", Proc. 12th Inter. 
Sympo. on Remote Sensing of Environment (ERIM), p.547-565 (1978) 
681 
ó Pd 
nci ain D ee s rem 
EMI 0 DE DAA Us UU QR IE Rn rm RE 
5 n " UA 
PME re EE i ee