Threats Detection in Aerial Objects: A Machine Learning Approach
View Polygence scholar page
The spy balloon that flew across the U.S. in February 2023, posed a serious security threat. US intelligence officials have said this balloon tried to gather intelligence by monitoring sensitive military sites and as a result the U.S. government began more closely scrutinizing its airspace to better categorize aerial objects and detect threats. However, the airspace is filled with a myriad of aerial objects, making the problem of classifying and risk determination very challenging. The purpose of this research is to develop a machine learning model to classify an aerial object based on the threat they pose. Currently, there are no known single datasets that contain both old and newer aerial objects such as drones, planes etc, nor datasets have labels to associate risk of the objects. The goal of this project is twofold: 1) We create a new comprehensive dataset that contains traditional & newer aerial objects. We label the objects into high, medium, low and no risk based on the threat markers on the aerial objects. 2) We use MobileNetV2 CNN classification algorithm to validate the dataset and provide accuracy results. Advancements in this space can potentially help intelligence agencies and security analysts to quickly assess developing scenarios and provide a reliable risk for observed aerial objects.
New Comprehensive Dataset of Aerial objects with novel idea of markers based labelling was created. MobileNetV2 Deep learning model was used to train on the new dataset. Test dataset of 40 unseen images were tested for prediction and 90% accuracy achieved on detecting risk on aerial objects
She helped me very much giving me feedback and support.