Robust object tracking for underwater robots by integrating stereo vision, inertial and magnetic sensors

As described in this paper, we propose a robust tracking system with a stereo vision, IMU, and magnetic sensor for underwater robots to estimate the position of a stationary object even when the object is out of the field of view (FOV) of a camera or occlusion occurs. During maneuvering of underwater robots, they are subject to external disturbances. Their cameras are always moving. Therefore, objects in an image move out of the FOV of a camera unexpectedly. As described in this paper, we integrate an IMU and a magnetic sensor with a stereo vision sensor to estimate an object’s position robustly. In the process of the estimation, Kalman filters are used to estimate the sensor system orientation. First, we conduct numerical analyses to investigate the estimation accuracy of the proposed method against sensor noise. Then, a preliminary experiment is conducted to demonstrate the robustness of the proposed method against occlusion and moving out of the FOV. As a first step, we examine the tracking performance with regard to rotational motion of the sensor system.

[1]  Suya You,et al.  A robust hybrid tracking system for outdoor augmented reality , 2004, IEEE Virtual Reality 2004.

[2]  Marie-Odile Berger,et al.  Use of inertial sensors to support video tracking , 2007, Comput. Animat. Virtual Worlds.

[3]  Junku Yuh,et al.  Underwater autonomous manipulation for intervention missions AUVs , 2009 .

[4]  Matthijs C. Dorst.  Distinctive Image Features from Scale-Invariant Keypoints , 2011 .

[5]  Andrew Zisserman,et al.  Multiple View Geometry in Computer Vision (2nd ed) , 2003 .

[6]  Luc Van Gool,et al.  SURF: Speeded Up Robust Features , 2006, ECCV.

[7]  Takeo Kanade,et al.  Gyro-aided feature tracking for a moving camera: fusion, auto-calibration and GPU implementation , 2011, Int. J. Robotics Res..

[8]  Greg Welch,et al.  Welch & Bishop , An Introduction to the Kalman Filter 2 1 The Discrete Kalman Filter In 1960 , 1994 .

[9]  Yuji Yamamoto,et al.  Observation support system of an ROV for underwater archaeology , 2015, 2015 International Conference on Intelligent Informatics and Biomedical Sciences (ICIIBMS).

[10]  Prashanth Swamy,et al.  An Improved Tracking using IMU and Vision Fusion for Mobile Augmented Reality Applications , 2014, ArXiv.

[11]  Takuji Nishimura,et al.  Mersenne twister: a 623-dimensionally equidistributed uniform pseudo-random number generator , 1998, TOMC.

[12]  Hideyuki Tamura,et al.  A hybrid registration method for outdoor augmented reality , 2001, Proceedings IEEE and ACM International Symposium on Augmented Reality.

[13]  Tom Drummond,et al.  Tightly integrated sensor fusion for robust visual tracking , 2004, Image Vis. Comput..

[14]  P ? ? ? ? ? ? ? % ? ? ? ? , 1991 .

[15]  Roland Siegwart,et al.  Fusion of IMU and Vision for Absolute Scale Estimation in Monocular SLAM , 2011, J. Intell. Robotic Syst..

[16]  Gary R. Bradski,et al.  ORB: An efficient alternative to SIFT or SURF , 2011, 2011 International Conference on Computer Vision.

[17]  Eric W. Frew,et al.  Relative position estimation for AUVs by fusing bearing and inertial rate sensor measurements , 2002, OCEANS '02 MTS/IEEE.

[18]  Angelo M. Sabatini,et al.  Extended Kalman Filter-Based Methods for Pose Estimation Using Visual, Inertial and Magnetic Sensors: Comparative Analysis and Performance Evaluation , 2013, Sensors.

[19]  Angelo M. Sabatini,et al.  Estimating Orientation Using Magnetic and Inertial Sensors and Different Sensor Fusion Approaches: Accuracy Assessment in Manual and Locomotion Tasks , 2014, Sensors.