Supervisors: Dr Zihao Wang, Dr KC Wong
Eligibility: WAM > 75, advanced understanding of computer vision, experience in Python, YOLO packages, data labelling
Project Description:
Bird strike is one of the most common aviation incidents that can sometimes lead to serious damage and accidents. The most recent example is the Jeju Air Flight 2216 that caused the loss of 175 passengers and 4 crews. Early warning about potential bird strikes can be issued to pilots if birds near the runway can be detected and monitored.
The project aims to investigate the system requirements for a bird detection system for a small general aviation aerodrome using computer vision model based on Flow Guided Feature Aggregation implemented on Pan-Tilt-Zoom cameras.
The student will be provided with an existing computer vision model and is expected to improve the model for bird and potentially other runway incursion. The student will also investigate the optimal camera positioning based on the field-of-view and range specifications. The goal is to validate the system at the Bankstown Airport.
Requirement to be on campus: Yes *dependent on government’s health advice.