Mapping of Trees outside Forest (ToF) From Sentinel-2 MSI Satellite Data using Object-Based Image Analysis 

Abstract

Trees outside Forests (ToF) have necessary socio-economic and environmental implications from local to global scales. In the current context of climate change, their importance is being increased dramatically for people's livelihoods and national economies, and also for various international processes that address global environmental and economic challenges like carbon sequestration, biodiversity loss, desertification, poverty alleviation etc. Assessment of Trees outside forests (TOF) is wide being recognized as an important theme, in sustainable natural resources management, because of their role in offering kind of goods, such as timber, fruits and fodder also as services like water, carbon, and biodiversity. Moderate resolution data (Sentinel-2A D) with different spatial resolution used, to get medium spectral and spatial resolution data (10m-60m range). Object-Based Image Analysis (OBIA) approach as implemented in commercial suite of e-Cognition (Version 9) consists of segmentation. The segmentation level of 1 O proved to be very precise in the geometric delineation of individual trees and groups of trees, which were the target groups in this research. An area of about 178.63 sq.km. was delineated from the satellite image using the segmentation approach and overall accuracy was found to be at 93.57%.The object-based approach provided by eCognition software is a big step forward in interpretations of remote sensing images and is an efficient and practical approach for information extraction.