Evaluation of the Species Composition and the Biological Productivity of Forests Based on Remote Sensing Data with High Spatial and Spectral Resolution

Abstract

The application of hyperspectral remote sensing of high spatial resolution is compared to conventional ground-based forest surveys on sample plots and is considered as a possible alternative to these labor-intensive works. Pattern recognition methods have become the principal approach used to solve this type of applied problems. Pattern recognition processing of hyperspectral images serves to identify different classes of objects as well as to determine their parameters, such as the net primary productivity of forests with different ages and species composition. The employed classifiers use the latest advances in forest pattern recognition based on hyperspectral images. The classification accuracy is compared to the accuracy of ground-based observations. The results indicate the promise of the proposed novel approach.