Document Type : Research Paper

Authors

• Fatemeh Tabib Mahmoudi
• Nahid Ghasemi

Shahid Rajaee Teacher Training University

Abstract

In the development of a country, natural resources and the environment are of great importance. Given the fact that Iran is one of the developing countries, and in terms of semi-desert vegetation, for sustainable development, we must pay special attention to the issue of the environment, including desert and desertification. Wilderness monitoring and desertification require extensive and continuous studies. Remote sensing technology can play a significant role in our studies due to the high regional use of satellite imagery along with temporal resolution and less cost, and time consuming than field study. In this thesis, satellite images of Landsat-7, ETM + and Landsat-8, OLI with time series were used in two time intervals of 15 years between 2002 and 2017. For each series, four images were taken with one month interval from each other, Of the months May, Jun, Julay and August from Qazvin plain area. After performing preprocessing on images, images were classified into four classes desert, vegetation, wet soil and otherobjedcts based on the spectral features and performing object based image analysis. After determining the classes for each of the four images per one year, classifier fusion was done based on the voting method. A final image was obtained from all four images in each year. Finally, detection of the changes between the two final classification results of 2002 and 2017 was performed, and the change map generated for desertification in the study area during these 15 years. Which reflects the desertification and rising deserts and the destruction of vegetation and soil moisture significant during these years in this plain.

Keywords

• Desertification
• Change detection
• Object Based Image Analysis
• Landsat Images