Comparison of Broadband and Hyperspectral Sensors for Lantana Mapping

Abstract

Remote sensing provides a useful tool for mapping invasive species across large areas. This study compared the effectiveness of imagery derived from three multi-spectral (Landsat Thematic Mapper (TM), SPOT 5, and Quickbird) and hyperspectral (Hyperion) imaging platforms for detection of the invasive species 'Lantana camara' L. in eastern Australia. Landsat TM multispectral (MS) image provided an overall accuracy of 85.1% and a kappa coefficient of 0.78 while SPOT 5 gave an overall accuracy of 84.9% and a kappa value of 0.77. Quickbird showed an overall accuracy of 84% and a kappa coefficient of 0.76. The Hyperion image gave an overall accuracy of 80% with a kappa value of 0.69. The highest producer (83%) and user accuracy (84%) for lantana was shown by Hyperion. The findings indicate that remote sensing technology can contribute to enhanced planning and decision making by scientists and resource managers involved in lantana research and management.