Classification of mature Australian pines (Casuarina equisetifolia) using LIDAR data in Broward County, Florida

The development of high resolution LIDAR DSM combined with digital infrared ortho-photography data enhances the ability to map canopy structures with a higher degree of accuracy and precision than with either data set alone. The purpose of this thesis is to map Australian Pine (Casuarina equisetifolia with a 85% or greater accuracy by creating a methodology that uses LIDAR and color infrared ortho-photography and to test it within three different landscape types within Broward County. LIDAR features below a determined height threshold (i.e. Deerpoint 25 ft) were eliminated and recoded to 0 to create Mask 1. NDVI technique separated non-vegetative features from vegetative features to create Mask 2. Mask 1 and Mask 2 were merged and overlaid on the raw LIDAR data set to perform isodata clustering, as well as density slicing to identify mature Australian Pines. Careful delineation of study areas is critical to obtain the highest possible accuracy. Density slicing proved to be a faster and less time consuming technique for achieving 85% level of accuracy than compared to isodata clustering.