Lab 10: Supervised Classification

Supervised Classification

This week we performed a supervised classification using tools in ERDAS Imagine. The final layout was created in ArcMap with the resulting Land Cover and Distance Images. The Distance Image is used to identify the brighter pixels that are more likely to have the wrong classification in the output image. Samples of spectral signatures were collected of known features and then used to create the Eight classes shown in the map. When applied properly, the supervised classification method produces a more accurate land cover map than the unsupervised method.