Fiche de projet de recherche

Onboard dynamic object detection and tracking for autonomous UAV

Programme d'études visé
Maîtrise avec projet

Domaines de recherche
Génie Électrique

According to the schedule, this Master’s student will be in charge of the following tasks:
1)    State-of-the-art review;
2)    Object detection algorithm design;
3)    Object tracking algorithm design;
4)    Object status estimation algorithm design;
5)    Module implementation and integration;
6)    Validation in lab environment;
7)    Final integration of real-time prototypes;
8)    Validation in real UAV system.
The general objective of the Master’s research project will be to research and implement object detection and tracking algorithm for autonomous UAV. Based on the image information, these algorithms will be run onboard to identify and track objects(both slow-moving and fast-moving), meanwhile provide the position, distance and speed information about the objects to the Flight Management System for mission planning and guidance. High accuracy and effectiveness of the algorithm will be great beneficial to the mission fulfillment of UAV systems and furthermore extend their application perspective. Several theoretical and technical aspects will be addressed, such as the computer vision, image processing, machine learning and software development.
For this purpose, the student will incorporate to the research team closely to fulfill the development and validation work .The student will firstly do the preliminary research work, review the state-of-the art in the target visual tracking method, and then propose several feasible technical solutions. Base on the system architecture and the software framework developed by other team members, the student will focus on the research work of real-time onboard dynamic target visual tracking. Several important algorithms such as Support Vector Machine (SVM) detector and particle filter would be researched in the application of object detection and tracking. The software module will be designed and implemented by the student. Relative experiments will be conducted both in the lab and outdoor environment to validate the accuracy, real-time capacity and robustness of the algorithm. As delivery, a software module package will be completed at the end of the project

Le projet débutera le 1er janvier 2018

scholarship: 15 000$ / year

Autres informations
Partenaires impliqués:
SiiCanada, VOZWIN, Concordia University, Polytechnique

Candidat recherché
Maîtrise avec projet
Good experience in embedded system and software development, expertise in image processing will be an asset.

Responsable à contacter
René Jr Landry  | Professeur|  Directeur du LASSENA|  Département de génie électrique
École de technologie supérieure  |  1100, rue Notre-Dame Ouest  |  Montréal (Qc) Canada  |  H3C 1K3
Bureau : A-2645  |  Tel.: 514 396-8506
Email :