Use of Remote Sensing Techniques in Lantana Mapping

Abstract

This study looked at the impact of image fusion as well as the cost effectiveness of using satellite imagery of varying spatial and spectral resolutions on the classification accuracy of lantana. The first part of the research compared the impact of four image fusion techniques using Quickbird imagery, namely Brovey, Hue-Saturation-Value, Principal Components and Gram−Schmidt Spectral Sharpening, on classification accuracies for lantana mapping. Gram−Schmidt and Principal Components spectral sharpening techniques showed an improvement in overall accuracy and the kappa coefficient compared to the MS image while Brovey transformation and HSV performed poorly in the supervised classification. The second part of the study compared the cost and classification accuracy of Quickbird, Landsat ETM+ and SPOT 5 imagery in mapping lantana. Quickbird multi-spectral image provided the highest overall accuracy followed by SPOT 5 image. Landsat ETM+ showed the lowest overall accuracy. SPOT 5 multispectral imagery provided a less costly option compared to Quickbird with no significant difference in overall accuracies between the two. The findings indicate that spatial resolution was of greater importance than spectral resolution in classifying lantana accurately and SPOT 5 provided the most cost-effective option.