Vision-inertial tracking algorithm with a known object's geometric model

Authors:
Ho Quoc Phuong Nguyen;Hee-Jun Kang;Young-Soo Suh
Affiliations:
Graduate School of Electrical Engineering, University of Ulsan, Ulsan, South Korea;School of Electrical Engineering, University of Ulsan, Ulsan, South Korea;School of Electrical Engineering, University of Ulsan, Ulsan, South Korea
Venue:
ICIC'10 Proceedings of the Advanced intelligent computing theories and applications, and 6th international conference on Intelligent computing
Year:
2010

Citing 3
Cited 0

Camera Calibration with Distortion Models and Accuracy Evaluation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Navigation Aided Image Processing in UAV Surveillance: Preliminary Results and Design of an Airborne Experimental System

Journal of Robotic Systems
An Introduction to Inertial and Visual Sensing

International Journal of Robotics Research

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper works toward the development and implementation of an INS/Vision integration algorithm for tracking an object with inertial measurement unit inside. For avoiding distortion and image processing error, an object's geometric model is used. For that purpose, Indirect Kalman filters are built as sensor fusion core. Experimental tests were performed with help of a mobile robot to hold and move the object in the view of stereo camera. Experiment result shows that object model which is embedded in process model helps improve the performance of the integration system prior to conventional one.