Investigating New Advances in Forest Species Classification

Detailed forest classification provides critical information for forest managers. The potential for species level classification from remotely sensed data has been challenging in the past because of limitations of both available image data and traditional classification techniques. Such limitations may be reduced by the increased availability of higher spatial resolution imagery as well as detailed digital elevation models e.g. derived from lidar collections. This project is a multi-year effort aimed at evaluating the benefits of combining traditional image classification techniques with derivatives of active remote sensing sources such as lidar for species level forest classification. The project plans to evaluate the relative benefits of different classification schemes, considering accuracy, efficiency, and effectiveness. The project focuses on the classification of imagery for the area in and around the Heiberg Memorial Forest in Tully, New York, utilizing existing forest inventory information for ground reference. The project aims at evaluating the applicability of different classification methodologies. Traditional approaches—such as supervised classification—provided a means to generate baseline classifications of satellite imagery (Landsat). This paper focuses on the classification of high spatial resolution QuickBird satellite imagery, with a goal of generating species level classification. Input to the classification includes a number of datasets derived from the imagery, as well as simple topographic characteristics derived from a digital elevation model. Incorporating such layers requires alternative methods of classification such as rule-based classifiers or neural networks. This project considers an object-oriented approach to rule-based classification. The analysis showed promising results for the separation of coniferous forest species. However, further research is needed to understand the benefits of different ancillary data layers as well as the derived data layers. INTRODUCTION AND OBJECTIVES