Robust pose estimation through visual/GNSS mixing
- 1 October 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 4, 1-8
Abstract
The current paper is dealing with the use of a novel system that takes advantage of several sensors (Global Navigation Satellite System-GNSS, Inertial Measurement Units-IMU and Digital Imaging Sensors - DIS) that are integrated on a single platform to visualize and provide augmented pictures of underground infrastructure networks (electricity, water, district heating, irrigation, sewerage and communication networks) in a city to help their better management and maintenance. A special tablet equipped with GNSS and IMU sensors theoretically is able to give precise information of the position (X, Y, Z) and attitude (ω,φ,κ) of the camera embedded on the system. The offset (from the GNSS antenna) of the centre of the lens and the rotation of the camera coordinate system can be determined through a system calibration procedure therefore the pose estimation parameters of the image taken from the Digital Imaging Sensor can be determined. In the case where the 3D GIS is available, that is mapping except from the city's roads, pavements and buildings the underground facilities networks, the image obtained by the camera of the system integrating the GNSS and IMU sensors can provide an augmented image of the facilities networks. This image will then be used for better management and maintenance purposes of the underground features in the city. There is always the possibility that the GNSS system is not working properly due to satellite invisibility and multi path error which is common during its operation in the city centre. Additionally, the magnetometer embedded in IMU is prone to environmental noise and interference and produces erroneous results. In this case two alternative methods are used to determine the exact exterior orientation of the camera: conventional photogrammetric pose estimation ; vanishing lines pose estimation All the methods are available in the system and depending on the GNSS signal quality and the adequacy of the control conditions the user can switch from one method to the others.Keywords
This publication has 6 references indexed in Scilit:
- RECORDING AND MODELLING OF MONUMENTS' INTERIOR SPACE USING RANGE AND OPTICAL SENSORSThe International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2016
- Sensor modelling and camera calibration for close-range photogrammetryISPRS Journal of Photogrammetry and Remote Sensing, 2016
- Inertial sensors for smartphones navigationSpringerPlus, 2015
- Calibration of Action Cameras for Photogrammetric PurposesSensors, 2014
- Autocalibration of MEMS AccelerometersIEEE Transactions on Instrumentation and Measurement, 2008
- Methods for in-field user calibration of an inertial measurement unit without external equipmentMeasurement Science and Technology, 2008