Change Detection of Olive Trees Distribution using Semi-Automated Object Based Image Classification

Rabia, Ahmed Harb; Abdelaty, Emad F; Elsayed, Maha Lotfy; Mohamed, Assem; Wassar, Fatma; Fiorillo, Edoardo; Di Vecchia, Andria; Tarchiani, Vieri

doi:10.21608/asejaiqjsae.2021.205324

Change Detection of Olive Trees Distribution using Semi-Automated Object Based Image Classification

Document Type : Original Article

Authors

¹ Department of Natural Resources and Agricultural Engineering, Faculty of Agriculture, Damanhour University, Damanhour, Egypt

² Damanhour University, Faculty of Agriculture, Department of Natural Resources and Agricultural Engineering, Alabaadia Assembl

³ Central Laboratory for Agricultural Climate (CLAC), Agricultural Research Centre, Giza, Egypt

⁴ Higher Institute of Water Sciences & Techniques, University of Gabès, Gabès, Tunisia

⁵ IBE-CNR, via Gobetti 101, 40129 Bologna, Italy

⁶ IBE-CNR, via dei Taurini 19, 00185 Rome, Italy

⁷ IBE-CNR, via Madonna del Piano 10, 50019 Sesto F.no, Florence, Italy

10.21608/asejaiqjsae.2021.205324

Abstract

Geographic object-based image analysis (GEOBIA) is a remote sensing technique that characterize image pixels into objects based on spectral, temporal, and spatial characteristics. It is a useful technique for land use classification and change detection. In this study, a land use and land cover classification and change detection was caried out at Oum Zessar watershed in the Medenine governorate of Tunisia to estimate the changes in olive trees distribution using high resolution satellite images of 2005 and 2013 and the geographic object-based image analysis technique (GEOBIA). Eight different vegetation indices (VIs) were used to enhance the classification process. The multi-resolution segmentation algorithm was selected as the main segmentation algorithm through the entire classification process. Results showed that Normalized Difference Vegetation Index (NDVI), Normalized Near Infrared (NNIR) and Ratio Vegetation Index (RVI) had high significance to be used for the recognition of the different objects and classes. In addition, results showed that olive tree canopy increased by almost 60% from 39 ha to 62 ha in the study area during the period from 2005 to 2013. In addition, analysis of the classification results showed that the number of the trees objects increased by 22.7 % from the year 2005 to 2013. This study showed the potential of Geographic object-based image analysis” (GEOBIA) technique in classifying land use in general and in detecting olive trees objects specifically.

Keywords