Vegetation Index-Based Classification of Pinus roxburghii and Quercus leucotricophora in Himalayan Temperate Forest Using Fuzzy Machine Learning Model

Accurate classification of tree species is essential for effective forest management, ecological monitoring and conservation efforts. In this study, we use Vegetation Indices (VIs) with the Individual Sample as Mean (ISM) training approach of Modified Possibilistic c-Means (MPCM) in semi-hypertemporal (SH) data for the classification of Pinus roxburghii and Quercus leucotricophora. We propose Modified Soil Adjusted Vegetation Index (MSAVI2) as the best VI among the selected three VIs for segregating the two target classes. Low variance, entropy and mean membership difference (MMD) values within the class show that the ISM training approach-based MPCM can handle heterogeneity within the target classes. This study highlights the potential of VI-based fuzzy modelling in the classification of tree species which is essential for sustainable forest management and conservation planning.