Accuracy comparison of Quickbird, Landsat TM and SPOT 5 imagery for Lantana camara mapping

Abstract

The distribution of an invasive species, Lantana camara L., was mapped using Landsat TM, SPOT 5 and Quickbird imagery with the maximum likelihood supervised classification method. The cost-effectiveness of the three different kinds of imagery was also evaluated. Landsat TM multi-spectral (MS) imagery provided an overall accuracy of 85.1 percent and a kappa coefficient of 0.78, while SPOT 5 imagery gave an overall accuracy of 84.9 percent and a kappa value of 0.77. Landsat TM was suitable for detecting dense infestations of lantana. Quickbird showed an overall accuracy of 84 percent and a kappa coefficient of 0.76. Spectral resolution and band position were more important than spatial resolution in classifying lantana accurately and the most cost-effective option was provided by Landsat TM with no significant difference in overall accuracies between the three types of imagery.

Keywords:

• Lantana (Lantana camara L.)
• invasive species
• supervised classification
• spatial resolution
• spectral resolution

Acknowledgement

The authors are grateful to Steven King from the Department of Environment, Climate Change and Water for providing access to the study area and support in field work and to Richard Coughlin for permission to use part of his dataset for this research. Thank you to Cate MacGregor for technical assistance. We would also like to thank the Department of Environment, Climate Change and Water for providing the SPOT 5 image.