GS2-3 A multithreaded algorithm of UAV visual localization based on a 3D model of environment: implementation with CUDA technology and CNN filtering of minor importance objects

Title:	GS2-3 A multithreaded algorithm of UAV visual localization based on a 3D model of environment: implementation with CUDA technology and CNN filtering of minor importance objects
Publication:	ICAROB2017
Volume:	22
Pages:	356-359
ISSN:	2188-7829
DOI:	10.5954/ICAROB.2017.GS2-3
Author(s):	Alexander Buyval, Mikhail Gavrilenkov, Evgeni Magid
Publication Date:	January 19, 2017
Keywords:	localization of UAV, particle filter, ROS, Gazebo, CUDA, CNN
Abstract:	Visual based navigation plays an important role in localization and path planning, especially in GPS-denied environments. This paper presents a visual based localization algorithm for a UAV within an indoor environment. The algorithm uses multithreaded computing CUDA technology and CNN-preprocessing filtering, which is responsible for filtering out dynamic objects. The algorithm is simulated in ROS/Gazebo environment with two different approaches – one uses CPU only and the other uses CPU and GPU - and their performance is compared.
PDF File:	https://alife-robotics.co.jp/members2017/icarob/data/html/data/GS_pdf/GS2/GS2-3.pdf
Copyright:	© The authors. This article is distributed under the terms of the Creative Commons Attribution License 4.0, which permits non-commercial use, distribution and reproduction in any medium, provided the original work is properly cited. See for details: https://creativecommons.org/licenses/by-nc/4.0/