Radiance 
Data Cube 
Remove Noise 
i 
Assess/Correct 
Data for Aircraft 
Motion Effects 
  
  
  
  
  
  
  
  
Roll Correction Y 
  
Retrieve Surface 
  
  
  
: 
  
  
Post-Process 
Cube 
  
  
Surface 
Reflectance 
   
Reflectance NS 
Select RT code 
Navigation 
Data 
RT Code 
Input Parameters 
  
  
  
  
  
  
Surface 
Reflectanc Cube 
Perform PCA d 
Select 
Endmember 
      
  
  
  
  
  
  
  
  
  
  
  
| d Unmix Cube á 
Spectra 
Crop 
Fractions 
Compute LAI 
  
  
  
     
  
  
  
  
  
  
  
  
  
  
Fi 1: Dat - ing flow (RT = Radiative Transfer). 
‘que Eat proces 00 adiiive Tima) Figure 3: Data processing layout for leaf area index (LAI) 
computation (PCA = Principal Component Analysis). 
60 
"asin sane casi reflectance 5 
50 - Smoothed casi reflectance 
à J 
< 404 
e 7 = [3:0 
3 Su t] a? o 
S E 
S 30 + 5 oo 
3 3 o o $ 
2 2 ER o 
i < 
3 | Bo o 
| lo L1 O Canola 
10 = O L] Wheat 
© Beans 
0 I Ï I Ï 0 T T T T 
$00 € 60050 (5700 8000800900‘ 1000 0 1 2 3 4 5 
Wavelength (nm) LAI (measured) 
  
Figure 2: casi surface reflectance of canola. The casi 
spectrum was averaged over four adjacent pixels. 
42 
Figure 4: Relationship between the leaf area index (LAI) derived 
from direct measurements LAI (measured) and from the 
image cube LAle (estimated). The solid line represents 
the x=y line. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998