ues 
as 
eld 
dera- 
from 
lows: 
assel- 
2403 (Panchromatic) 
2h24 (Infrared Black-and-White), 
with suitable filters. 
Scale of photography : 1 2 20,000: 
Each frame (effective area = 55mm x 55mm) represents an 
area of 1,l1km x 1.1km (approximately) on the ground. 
The period suitable for an aerial survev in the area under 
study is (end of)October - (beginning of)April. The main data 
were acquired in the months of November and March, and selective 
data in the months of October and December of the year subse- 
quent to the main flights. 
In view of some nonuniformity (with respect to time) of the 
sowing and harvesting practices, one cannot expect, at any par- 
ticular time of the year, the same growth stage of the crop in 
the whole area. However, the growth stages for paddy and sugar- 
cane seem to be ôf the fotlowing general pattern 
Growth Stages of 
Paddy Sugarcane 
November March November March 
A1 .3 A1.1 A2.2 A243 
Al. A2.h A2. 
Test sites were chosen in randomly located places for col- 
lecting the so-called "ground truth" data synchronous with the 
flight operations over the area. This data serves as the train- 
ing set for both visual and automatic data processing. 
Remark In an interesting comparative study of the various re- 
mote sensors by H.W.Faust (Carl Zeiss. Oberkochen), 37th Photo- 
grammetric Week (Stuttgart, September 1977), it is convincingly 
brought out that Hasselblad cameras are the cheapest means of 
acquiring multiband imagery (in the visible and near infrared 
portions of the electromagnetic spectrum) with exceptionally 
good resolution. For instance, assuming 50 lines / mm resolu- 
tion in the aerisl film. each pixel represents O.lhm x O.hm (ap- 
proximately) on the ground, at the 1:20,000 seale. 
5. Processing of Photographi c Data An automatic processing 
system should at least be able to do whatever the visual system 
can do. This requires that we should know exactly what the human 
visual system can do. 
With respect to the project, the color infrared frames 
have been visuallv internreted by trained photointervoreters.Se- 
lected frames have been checked against ground data. Standard 
techniques are employed for calculation of acreage of each cate- 
gory. Yield calculations are made on the basis of average vield 
per acre. 
Visual interpretation is based on color, texture and con- 
textual information on the color infrared frame. It was found 
difficult to interpret panchromatic and infrared black-and-white 
films. It has been reported in the literature that a color infra 
red frame contains more information on agricultural and nonagri-