Estimation Methods for Fractional Vegetation Cover Using UAV Remote Sensing

摘要

Abstract

Fractional vegetation cover is an important parameter for evaluating ecological conditions and crop growth. This paper analyzes the technical workflow and principal methods for estimating fractional vegetation cover from unmanned aerial vehicle (UAV) RGB and multispectral imagery. The findings indicate that threshold segmentation and the two-endmember pixel model are straightforward to implement, whereas spectral unmixing and machine learning are better suited to complex backgrounds. However, estimation results are affected by illumination, shadows, phenology, endmember selection, and spatial scale. Future research should standardize the definition of fractional vegetation cover and its validation protocols, strengthen multitemporal and multisensor data fusion and cross-regional generalization, and establish a traceable uncertainty-assessment system.