Please use this identifier to cite or link to this item:
Title: Real-time drone classification and detection using deep learning
Authors: Thadani, Jashan Vishindas
Keywords: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
Engineering::Aeronautical engineering
Issue Date: 2021
Publisher: Nanyang Technological University
Source: Thadani, J. V. (2021). Real-time drone classification and detection using deep learning. Final Year Project (FYP), Nanyang Technological University, Singapore.
Project: C071
Abstract: This report is written to document and sum up all the findings and research of the Final Year Project. The aim is to achieve real-time drone detection with a static camera. The method used in this research project is utilizing deep learning methods in computer vision where convolutional neural networks are primarily used to train the built model over an appropriate dataset such that after training, it would be able to work in real-time on new pieces of data input into it. The models built consists of two units: Classification and Detection, in which they both work in tandem to achieve working results. The classifier looks at a portion of given frame or image and determines whether it is a drone image or not and the detector looks a the entire frame to determine which region the drone image is likely to be in. The objectives were met by the end of this final year project and some decent results were obtained. In real-time, the model and classifier can detect a drone in an untested video footage with a high level of accuracy and relatively quick speed. The limitations of the model built is that it requires a powerful computing device to achieve the same results in higher frames per second.
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:MAE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
Final Report Jashan Final-converted.pdf
  Restricted Access
FYP Final Report - Jashan1.45 MBAdobe PDFView/Open

Page view(s)

Updated on Nov 29, 2021


Updated on Nov 29, 2021

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.