Precise input and output error characterization for loosely integrated INS/GPS/camera navigation system

Lee Lemay, Chen Chi Chu, Demoz Gebre Egziabher, Rohan Ramllal

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Scopus citations

Abstract

A method for the loose integration of INS, GNSS and cameras using an Extended Kalman Filter (EKF) is presented. The integration is loose because it occurs at the level of position, velocity and attitude. The GNSS supplies position and velocity estimates (when available) and the camera provides position and attitude estimates. The information provided by GNSS and the camera are used in the EKF to estimate inertial sensor errors and, thus, mitigate INS drift. Since integration via an EKF requires accurate characterization of errors (used as weights in the calculation of gains), a method for estimating camera position and attitude covariance as function of pixel noise is presented. A series of Monte Carlo simulations show that the proposed approach estimates the camera position covariance well but underestimates the attitude covariance. Simulation results showing the performance of an integrated INS/GNSS/Camera system using the developed covariance estimator is presented. The results show that an integration with 1 Hz camera position/attitude updates is sufficient for estimating the inertial sensor errors of an INS which uses an automotive grade IMU.

Original languageEnglish (US)
Title of host publicationInstitute of Navigation - International Technical Meeting 2011, ITM 2011
Pages880-894
Number of pages15
StatePublished - May 26 2011
EventInstitute of Navigation - International Technical Meeting 2011, ITM 2011 - San Diego, CA, United States
Duration: Jan 24 2011Jan 26 2011

Publication series

NameInstitute of Navigation - International Technical Meeting 2011, ITM 2011
Volume2

Other

OtherInstitute of Navigation - International Technical Meeting 2011, ITM 2011
CountryUnited States
CitySan Diego, CA
Period1/24/111/26/11

Fingerprint Dive into the research topics of 'Precise input and output error characterization for loosely integrated INS/GPS/camera navigation system'. Together they form a unique fingerprint.

Cite this