Supawan Wongprayoon 1, Carlos A.O. Vieira 2, and Joseph J. H. Leach 1
1 Department of Geomatics, 
The University of Melbourne, 
Parkville , VIC 3010,Australia 
s.wongprayoon@pgrad.unimelb.edu.au, leach@unimelb.edu.au
2 Departamento de Engenharia Civil, 
Universidade Federal de Viçosa, 
Viçosa, MG, 36570-000, Brazil 
carlos.vieira@ufv.br

Abstract
This paper describes methods of assessment of classifier performance for coral reef classifications that explicitly include the spatial pattern of classification errors, and which presents the user with a visual indication of the reliability of the pixel label assignments. Two coral reef test sites in Thailand were used in this study. Two of Ikonos images were also used covering the respective sites.  It was performed a supervised classification using the maximum likelihood classifier. Non-spatial statistics such as: overall accuracy, Kappa coefficient, variance, and Z statistics; were computed from the error matrix and thematic images were also generated.  None of these statistics explicitly considers the spatial distribution of misclassified pixels. This paper describes methods of assessment of classifier performance for coral reef classifications that explicitly include the spatial pattern of classification errors, and which presents the user with a visual indication of the reliability of the pixel label assignments. Results show that despite statistical accuracy measurements which present outstanding results (values), the accuracy is not spatially  similar for every pixel through the image (homogeneous).  Thus, a considerable amount of research and development needs to be accomplished before the spatial characterization of thematic accuracy associated with remote sensing products can be adequately reported in standardized format and legends.

Keywords: reliability of the remote sensing products, thematic accuracy, coral reef mapping, IKONOS, image classification

In: Caetano, M. and Painho, M. (eds). Proceedings of the 7th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences, 5 – 7 July 2006, Lisboa, Instituto Geográfico Português