Date Approved


Embargo Period


Document Type


Degree Name

M.S. Electrical and Computer Engineering


Electrical and Computer Engineering


Henry M. Rowan College of Engineering


Nidhal Bouaynaya, Ph.D.

Committee Member 1

Charles C. Johnson

Committee Member 2

Gregory Ditzler, Ph.D.


image segmentation, object detection, runway identification


Runways (Aeronautics); Deep learning (Machine learning)


Aerospace Engineering | Electrical and Computer Engineering


The United States lacks a comprehensive national database of private Prior Permission Required (PPR) airports. The primary reason such a database does not exist is that there are no federal regulatory obligations for these facilities to have their information re-evaluated or updated by the Federal Aviation Administration (FAA) or the local state Department of Transportation (DOT) once the data has been entered into the system. The often outdated and incorrect information about landing sites presents a serious risk factor in aviation safety. In this thesis, we present a machine learning approach for detecting airport landing sites from Google Earth satellite imagery. The approach presented in this thesis plays a crucial role in confirming the FAA's current database and improving aviation safety in the United States. Specifically, we designed, implemented, and evaluated object detection and segmentation techniques for identifying and segmenting the regions of interest in image data. The in-house dataset has been thoroughly annotated that includes 400 satellite images with a total of 700 instances of runways. The images - acquired via Google Maps static API - are 3000x3000 pixels in size. The models were trained using two distinct backbones on a Mask R-CNN architecture: ResNet101, and ResneXt101, and obtained the highest average precision score @0.75 with ResNet-101 at 92% and recall at 89%. We finally hosted the model in the StreamLit front-end platform, allowing users to enter any location to check and confirm the presence of a runway.