No.
2.1)
) 2
Table 1. Two category classification using the three band data
Remarks i)
ii!
iii)
iv)
v)
Table 2.
No.
+
2
srovolts) 3
of the h
as shown
Categories % Correct No. Categories % Correct
Classification Classification
AYX.l 100 7 Al.3 88
A?2.1 0 A2.1 60
A1.1 82 8 A1.3 0
A2.2 51 A2.2 96
Al. À 97 9 Al.3 90 .
A2.3 0 A2.3 18 (ie. 6)
A1.2 92 10 A34 95 :
Abl 90 A2.1 oa (Fie.T)
AI .? 20 13 A1. 96
A2.2 95 A2.2 95
An 2 93 12 A1.4 96
42.3 95 A2.3 9h
All the pixels of sugarcane freshly planted (A2.1)
were misclassified as paddy freshly planted (A3.1).
However, none of the pixels of paddy freshlv planted
were misclassified as sugarcane freshly planted.
Similarly for paddy freshly planted (A1.1) and sugar-
cane young ratoon (A2.2) ; and also for paddy freshly
planted and sugarcane well grown (42.3).
In the area chosen for digital analysis, at the time
of taking aerial photographs,sugarcane had not reach-
ed the mature stage/not been harvested. Hence no pix-
els of A2.h and A2.5. ;
The pair, paddy well grown (A1.2) and sugarcane well
grown (A2.3), among a few other pairs, can be classi-
fied with sufficient accuracy using the three band
information of the color infrared frame.
Note that various stages of growth of both paddy and
sugarcane can appear even in a small area.
Choice of band pairs for the classification of mature
paddy (A1.3) and sugarcane well grown (A2.3}).
Bands chosen % Accuracy of classification
AY.3 A243
Blue and green 46 31
Red and green 40 68
Red and blue 55 86
Blue, green and red 90 78