Detection and Tracking in Airborne Image Sequences

26 de junho de 2018

This work proposes and evaluates a method for detection and tracking of maritime vessels in airborne image sequences. Such sequences are challenging due to sun reflections, low resolution, wakes, wave crests and fast motions either from the vessel but also from the UAV (Unmanned Aerial Vehicle), which significantly degrade the performance of general purpose tracking algorithms. The proposed method is based on state-of-the-art deep neural network detection method complemented with a correlation filter tracker. We evaluate our proposal using a known benchmark in the field and compare the obtained results with the results obtained with the original algorithms. The dataset used to perform the evaluations was obtained during the SEAGULL project.

Autores da comunidade :

• 

  P

  Patrícia Maria Gonçalves Silva

  ist170411

Orientadores desta instituição:

• 

  J

  Jorge Dos Santos Salvador Marques

  ist11994

• 

  A

  Alexandre José Malheiro Bernardino

  ist13761

mechanical-engineering - Engenharia Mecânica

por - Português

Instituto Superior Técnico