Adaptive Object-Oriented and Spectral-Based Geo-Processing for Tree species Mapping using Sentinel-2A and EO-1 Hyperion Hyperspectral Satellite Imagery in FRI Campus

Accurate mapping of tree species is essential for effective forest management and conservation. This paper explores various methods for tree species mapping, focusing on the Forest Research Institute (FRI). Authors evaluate GIS and remote sensing techniques through a case study within the FRI campus, presenting an innovative approach that integrates adaptive object-oriented and spectral-based methods. Utilizing moderate resolution Sentinel-2A and hyperspectral EO-1 Hyperion data enhances forest type classification precision. The methodology segments satellite imagery into meaningful objects and applies spectral analysis, achieving an overall classification accuracy of 92.5% and a Kappa coefficient of 0.89. The findings of the study demonstrate the effectiveness of this approach over traditional methods, providing valuable insights for future forest monitoring and management strategies.