te 
ne 
ed 
ed 
ag 
ne 
he 
S. 
1ts 
Ebrahim, Mostafa Abdel-Bary 
  
2 Sa) 
about 350 km” of dry valley that has a special importance for the future planning of Assiut Governorate. In the central 
part of this area some private agricultural activities and new communities were started and planned to be extended 
[Farrag (1997)]. 
Three SPOT multispectral images with 20 m ground resolution were the main data source for this study. The SPOT 
images correspond to column 113 and row 296 of SPOT Grid Reference System (GRS). The first image was captured 
by SPOT-1 on 23 July 1992, the second image was captured by SPOT-3 on 25 July 1995, and the third image was 
captured by SPOT-4 on 11 August 1998. SPOT-1, 2, 3 and 4 are equipped with High-Resolution Visible (HRV) sensor 
that recording three bands in the multispectral mode with 20 m ground resolution. Band 1 (the green band) covering the 
spectral band from 500 nm to 590 nm, band 2 (the red band) covering the spectral band from 610 nm to 680 nm , and 
band 3 (the near infra red band) covering the spectral band from 790 nm to 890 nm [Michele 1981 and SPOT Magazine 
No 28 ( January 1998)]. 
3. METHODLIOLOGUY 
The process and analysis of SPOT images consisted of four steps. First, the images were geometrically corrected. 
Second, the data of each of the three bands was examined and analyzed individually. Third vegetation indices were 
constructed and examined. Fourth, a compost image was constructed and examined visually and unsupervised 
classification was carried out in order to test the possibility of extracting the reclaimed areas automatically. Geometric 
correction was applied to the part of the image which cover the two study areas, while the process of examining the 
information contents of individual image bands, examining vegetation indices, and unsupervised classification were 
applied to the test area. Then the processing technique which give most accurate results was applied to the second study 
area. 
3.1 Geometric Correction 
Geometric correction of SPOT images was essential in order to be able to take measurements from the 
images and to carry out quantitative analysis. The images were rectified to the Universal Transverse 
Mercator (UTM) coordinate system using a second order polynomial transformation. Each of the three bands 
of each image was individually geometrically corrected using the same set of control points. The measurement of the 
ground coordinates of the control points was carried out in the field applying Differential Global Positioning 
System (DGPS). The residual Root Mean Square (RMS) error was 20.1 m (about one pixel). The resampling process 
of each band was carried out using nearest neighbor method (in order to avoid any smoothing to the original image 
data). 
32 Image Stretching 
In this step major emphasis was made on the examining the sensitivity of each of the three bands of SPOT 
image to detect the new reclaimed area. The identification and delineation of new reclaimed areas were 
based on the information contents of individual bands. 
The image of each of the three bands was processed separately and the obtained results were compared. 
Image processing includes linear stretching and slicing images to a number of levels. The process of linear 
stretching involves identification of two brightness values (DNs) as minimum and maximum values in the resulting 
stretched image. These minimum and maximum values represent the limits of the feature of interest (the new 
reclaimed area in our case). Then the image data between the two limits was sliced to a number of levels, to be 
interpreted as different classes of reclaimed areas. The determination of the minimum and maximum values that 
corresponding to the reclaimed land in the test area was carried out by displaying the image of each band on the screen 
and testing the pixels that corresponding to the well defined reclaimed areas (the test area). The identified values are 
given in Table 1. 
In general, it realized that band 3 (the near infrared band) of SPOT images is the most sensitive band to reclaimed and 
vegetated areas. In order to include the information that can be obtained from the other two bands, composite images 
were constructed from the stretched bands and given for the test area in Plates 1, 2, and 3 for the images of 92, 95, and 
98 respectively. 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 383