Deep learning based tools for drone surveillance and detection in adverse environmental conditions

• Main Author: Chia, Wei Fong
• Other Authors: Lye Sun Woh
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Aeronautical engineering::Accidents and air safety
• Online Access: https://hdl.handle.net/10356/159074
• Institution: Nanyang Technological University

This study explores the design of an automated Machine pipeline comprising state-of-theart image enhancement and object detection algorithms as an aid for air traffic controllers to quickly spot and identify drone incursions in the surrounding airspace. Experiments were conducted to evaluate the drone detection performance of the Machine pipeline by itself, of human capabilities by themselves and a Human-Machine collaboration with human operators and the Machine pipeline. Results suggest that by human effort alone, drone detectable range and spatial awareness were lacking for effective detection. By machine effort alone, presence of errors limits use of the Machine pipeline in actual air traffic management where there is a low tolerance for errors for safety reasons. Rather, a Human-Machine collaboration is shown to be optimal as the Human and Machine components compensate for each other’s shortcomings while complementing in strong points, leading to improved drone detection performance.