| This study investigates an alternative classification algorithm, NEFCLASS, and its ability to classify remote sensing images. NEFCLASS is a Neuro-fuzzy System that is capable of generating a set of linguistic rules. These rules allow the user to check and interpret the classification results. This study also shows that the neural net rules stabilized after only a few training iterations. The land-use/land-cover classification result produced by NEFCLASS is compared to the result produced by a conventional classification algorithm, Maximum Likelihood Classifier (MLC). NEFCLASS produced better classification accuracy than MLC. |
