USING VISUAL TECHNIQUES TO DETERMINE THE CHANGES USING VISUAL TECHNIQUES TO DETERMINE THE CHANGES

Document Type : Original Article

Authors

1 Computer Science Dep., Faculty of Computer and Information Sciences, Ain Shams University- Egypt

2 Misr Higher Institute for Commerce and Computer Science, Mansoura-Egypt

3 Computer Science Dept., Faculty of Computer and Information System, Mansoura University - Egypt

Abstract

Building new cities at the fringes of old ones is a mandatory nowadays to lower the over
increasing population in old cities, and to decrease the heavy load on the infrastructure and services.
The main objective of this work was to evaluate the spatial and temporal changes in land uses within
the studied area by using Remote Sensing (RS) and Geographic Information Systems (GIS) data and
techniques. This is in addition to providing accurate estimations of current land uses to support
decision makers with the right information for further development. Accordingly, Landsat TM images in
1984 and 1999 and Landsat 8 in 2014 were used in this study. Normalized difference vegetation
difference index (NDVI) was used to map agricultural versus non - agricultural lands. Also, the
modified normalized difference water index (MNDWI) was used to map dry lands versus wet lands
(Fish pounds) in the area. The obtained results indicated that agricultural lands were increased by
about 23.1 km2 from 1984 to 1999 and by about 30.1 km2 from 1999 to 2014. The total increase in
agricultural lands in 30 years from 1984 to 2014 was about 53.2 km2. That increase in agricultural
lands was due to land reclamation projects north of Nile-Delta. On the other hand, water features were
increased by about 16.3 km2 from 1984 to 1999 and by about 23.0 km2 from 1999 to 2014. The total
increase in Water features from 1984 to 2014 was about 39.3 km2. That increase in water features was
mainly due to the development of fish pounds. Land use classification derived from the gap-filled
Landsat SCL-off image acquired in 2009 was more accurate when the gap-filling was carried out by
using the Landsat gap-fill plug-in ENVI than using the Matlab. The overall accuracy of the gap-filled
images was not very high, where the gap-filling algorithms could not retrieve the actual pixel values but
interpolate them.

Keywords

Files

Share

How to cite

Statistics