Deep learning based on LSTM model for enhanced visual odometry navigation system

被引:10
作者
Deraz, Ashraf A. [1 ,2 ]
Badawy, Osama [3 ]
Elhosseini, Mostafa A. [4 ,5 ]
Mostafa, Mostafa [6 ,7 ]
Ali, Hesham A. [4 ,8 ]
El-Desouky, Ali I. [4 ]
机构
[1] Air Def Coll, Dept Res, Alexandria, Egypt
[2] Mil Tech Coll, Dept Comp Engn & Operat Res, Cairo, Egypt
[3] Arab Acad Sci Technol & Maritime Transport, Coll Comp & Informat Technol, Alexandria, Egypt
[4] Mansoura Univ, Fac Engn, Comp & Control Syst Engn Dept, Mansoura 35516, Egypt
[5] Taibah Univ, Coll Comp Sci & Engn, Yanbu 46421, Saudi Arabia
[6] Univ Calgary, Dept Geomatics Engn, Calgary, AB, Canada
[7] Air Def Coll, Dept Elect & Elect Commun Engn, Alexandria, Egypt
[8] Delta Univ Sci & Technol, Fac Artificial Intelligence, Mansoura, Egypt
关键词
Deep learning; Long Short-Term Memory Networks; Multi-sensor data fusion; INS; VO; INS/GPS INTEGRATION; NEURAL-NETWORKS; RADAR ODOMETRY; KALMAN FILTER; GPS; REGRESSION; METHODOLOGY; ALGORITHM; VEHICLES; HYBRID;
D O I
10.1016/j.asej.2022.102050
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
UAVs are employed for military, commercial, environmental, and other objectives. Flying in complex situations might strain the GPS (GNSS). Using INS alone increases positional inaccuracy. Despite cameras and sensors, drift persists. This work provides a GNSS-free UAV navigation system employing optical odometry, radar height estimates, and multi-sensory data fusion. Our monocular VO with optical flow leverages LSTM networks. We use optical flow to determine the vehicle's forward speed and LSTM to correct drift. A five-set LSTM model trained on GNSS data produces the velocity difference. The suggested technology was flight-tested. Experiments indicate the system can counteract lost GNSS signals' effects on forward and lateral speed. When GNSS signals are lost, the proposed strategy reduces average forward and lateral velocity errors to 63.01% in 30 s, 62.26% in 60 s, 58.76% in 90 s, and 54.33% in 113 sec. (c) 2022 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Ain Shams University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/ by-nc-nd/4.0/).
引用
收藏
页数:13
相关论文
共 70 条
[1]  
Abdolkarimi ES, 2015, 2015 SIGNAL PROCESSING AND INTELLIGENT SYSTEMS CONFERENCE (SPIS), P93, DOI 10.1109/SPIS.2015.7422319
[2]  
Achtelik Markus, 2011, IEEE International Conference on Robotics and Automation, P3056
[3]   A novel hybrid approach utilizing principal component regression and random forest regression to bridge the period of GPS outages [J].
Adusumilli, Srujana ;
Bhatt, Deepak ;
Wang, Hong ;
Devabhaktuni, Vijay ;
Bhattacharya, Prabir .
NEUROCOMPUTING, 2015, 166 :185-192
[4]   A low-cost INS/GPS integration methodology based on random forest regression [J].
Adusumilli, Srujana ;
Bhatt, Deepak ;
Wang, Hong ;
Bhattacharya, Prabir ;
Devabhaktuni, Vijay .
EXPERT SYSTEMS WITH APPLICATIONS, 2013, 40 (11) :4653-4659
[5]   Obstacle Detection and Avoidance System Based on Monocular Camera and Size Expansion Algorithm for UAVs [J].
Al-Kaff, Abdulla ;
Garcia, Fernando ;
Martin, David ;
De la Escalera, Arturo ;
Maria Armingol, Jose .
SENSORS, 2017, 17 (05)
[6]  
ARTECH HOUSE USA, DESIGN ANAL MODERN T
[7]   Design of Low Altitude Long Endurance Solar-Powered UAV Using Genetic Algorithm [J].
Bakar, Abu ;
Ke, Li ;
Liu, Haobo ;
Xu, Ziqi ;
Wen, Dongsheng .
AEROSPACE, 2021, 8 (08)
[8]  
Banerjee S., 2014, LINEAR ALGEBRA MATRI
[9]   Speeded-Up Robust Features (SURF) [J].
Bay, Herbert ;
Ess, Andreas ;
Tuytelaars, Tinne ;
Van Gool, Luc .
COMPUTER VISION AND IMAGE UNDERSTANDING, 2008, 110 (03) :346-359
[10]  
Bryson M, 2009, IEEE INT CONF ROBOT, P3143