Identification of Forest Vegetation Using Vegetation Indices

Abstract

Spectral feature of forest vegetation with remote sensing techniques is the research topic all over the world, because forest plays an important role in human beings' living environment. Research on vegetation classification with vegetation index is still very little recently. This paper proposes a method of identifying forest types based on vegetation indices, because the contrast of absorbing red waveband with reflecting near-infrared waveband strongly for different vegetation types is recognized as the theoretic basis of vegetation analysis with remote sensing. Vegetation index is highly related to leaf area index, absorbed photosynthetically active radiation and vegetation cover. Vegetation index reflects photosynthesis intensity of plants and manifests different forest types. According to reflectance data of forest canopy and soil line equation NIR=1.506R+0.0076 in Jingyuetan, Changchun of China, many vegetation indices are calculated and analyzed. The result shows that the relationships between vegetation indices and forest types are that perpendicular vegetation index (PVI) identifies broadleaf forest and coniferous forest the most easily; the next is transformed soil-adjusted vegetation index (TSVI) and modified soil-adjusted vegetation index (MSVI), but their calculation is complex. Ratio vegetation index (RVI) values of different coniferous forest vary obviously, so RVI can classify conifers. Therefore, the combination of PVI and RVI is evaluated to classify different vegetation types.