Skip to main content

A Comparison of Machine Learning Techniques to Extract Human Settlements from High Resolution Imagery...

by Jeanette E Weaver, Brian M Moore, Andrew Ej Reith, Jacob J Mckee, Wadzanai D Lunga
Publication Type
Conference Paper
Journal Name
IEEE International Geoscience and Remote Sensing Symposium (IGARSS)
Publication Date
Page Numbers
6412 to 6415
Conference Name
2018 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2018)
Conference Location
Valencia, Spain
Conference Sponsor
The Bill and Melinda Gates Foundation
Conference Date

Two machine learning techniques were developed to extract human settlements from very high resolution (VHR) satellite images of 3 provinces in Afghanistan: Logar, Panjsher, and Wardak. The results were then compared with analyst verified reference data information known as the LandScan Settlement Layer (LandScan SL).[1] This study attempts to compare settlement mapping results from a support vector machine (SVM) classifier specifically integrated in a current settlement mapping framework and a deep learner utilizing a convolutional neural network (CNN) approach. By comparing the results from the SVM and the CNN to the reference data information we demonstrate that the CNN yields more accurate results overall, in terms of overall pixel cells, and the SVM performs more accurately in omission, based on derived statistics against the reference data information.