Page 703 - NGTU_paper_withoutVideo

Modern Geomatics Technologies and Applications

In this study, both methods have been used to detect changes. The change and non-change classes are displayed in a
general change map and also the type of classes’ changes to another is specified.

4. Experimental Results
In this section, the results of implementation of the proposed change detection method on multi-temporal Landsat
satellite images taken from the new city of Pardis are presented and the results will be explained in detail. In the following,
considering a relation between the time and stages of construction of different phases of the Mehr Pardis housing project and
land use/ cover changes, the rate and manner of these changes will be discussed and evaluated.
The first step in applying pre-processing on multi-temporal satellite images was to perform atmospheric correction. In
this research, the FLAASH atmospheric correction module of ENVI 5.3 software was used on all image bands. Then, for
sharpening the spectral bands of Landsat images and increase their spatial resolution to 15 meters, the Gram-Schmidt pan-
sharpening method in ENVI 5.3 software was used to produce a pan-sharpened product from the panchromatic and multi-
spectral satellite images. Therefore, due to the fact that both Landsat-7 ETM + and Landsat-8 OLI sensors have a panchromatic
band with a spatial resolution of 15 meters, the produced pan-sharpened products also have spectral bands with a spatial
resolution of 15 meters. Fig. 3 shows the result of applying the pre-processing step to the multi-temporal images.

a b
Fig. 3. The results of performing pre-processing on a)Landsat-7 ETM+ image taken in 2002,
b) Landsat-8 OLI image taken in 2019

For object based classification of the images, first the multi-resolution segmentation algorithm is applied in eCognition
software environment. In multi-resolution segmentation of the Landsat-7 ETM+ image of 2002, the values 1, 0.5 and 0.5 are

a b
Fig. 4. Knowledge based classification maps of a)Landsat-7 ETM+ image taken in
2002, b) Landsat-8 OLI image taken in 2019

4