Joel Sexson's GIS Blog UWF

    Hello all, in this week of  Photo Interpretation and Remote Sensing, we learned how digital imagery is classified both unsupervised and supervised. Unsupervised classification is done manually and does not use a predetermined classification that matches pixels. Supervised classification is done under the supervision of an already established set of "spectral signatures". We used ERDAS Imagine to classify the images and ArcGIS to make the layout.  This image uses a band combination that helps each feature stand out and avoid being spectrally confused. The main image shows the classifications, explained in the legend on the side. The distance map shows the accuracy of the classification. Brighter spots on the distance map were more likely to have been spectrally confused.  Thank you,  Joel Sexson

 Hello all, in module 2 of my PI&RS course, we practiced land use/land classification (LULC) on a photo. We also got practice with digitally truthing to test out the accuracy of our classifications.  I have included the map from the exercise below. The map shows the LULC in slightly transparent colors and the accuracy of the random points. The legend gives a brief description of each type of LULC code. This exercise was useful in learning how to use some interpretation skills that I learned from the previous model on classifying areas within a photo.  Thank you, Joel Sexson