Skip to content

Gaussian Mixture Model and Rule Based Classifier based Settlement Mapping.

Notifications You must be signed in to change notification settings

amwat/GMM-RBC-Settlement-Mapping

Repository files navigation

GMM + RBC for Settlement Mapping

A combination of unsupervised learning (soft clustering through a Gaussian Mixture Model) and supervised learning (classification via a Rule-Based Classifier using the GMM results as attributes) is used to automate the process of classifying patches of satellite images as one of five area types: commercial, residential 1, residential 2, water, and vegetation. Both the GMM and RBC were built from scratch. The final results of our Rule-Based Classifier are of comparable quality to the Rule-Based Classifier used in the Weka package.

We have written a paper on this project and also have created a poster.

Poster

About

Gaussian Mixture Model and Rule Based Classifier based Settlement Mapping.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published