Robotics Vision-based Heuristic Reasoning for Underwater Target Tracking and Navigation

This paper presents a robotics vision-based heuristic reasoning system for underwater target tracking and navigation. This system is introduced to improve the level of automation of underwater Remote Operated Vehicles (ROVs) operations. A prototype which combines computer vision with an underwater robotics system is successfully designed and developed to perform target tracking and intelligent navigation. This study focuses on developing image processing algorithms and fuzzy inference system for the analysis of the terrain. The vision system developed is capable of interpreting underwater scene by extracting subjective uncertainties of the object of interest. Subjective uncertainties are further processed as multiple inputs of a fuzzy inference system that is capable of making crisp decisions concerning where to navigate. The important part of the image analysis is morphological filtering. The applications focus on binary images with the extension of gray-level concepts. An open-loop fuzzy control system is developed for classifying the traverse of terrain. The great achievement is the system's capability to recognize and perform target tracking of the object of interest (pipeline) in perspective view based on perceived condition. The effectiveness of this approach is demonstrated by computer and prototype simulations. This work is originated from the desire to develop robotics vision system with the ability to mimic the human expert's judgement and reasoning when maneuvering ROV in the traverse of the underwater terrain.

[1]  A. R. Greig,et al.  Robust scene interpretation of underwater image sequences , 1997 .

[2]  M. Zampato,et al.  Segmentation of underwater images for AUV navigation , 2000, Proceedings of the 2000. IEEE International Conference on Control Applications. Conference Proceedings (Cat. No.00CH37162).

[3]  F. El-Hawary,et al.  A robust pre-filtering approach to EKF underwater target tracking , 1993, Proceedings of OCEANS '93.

[4]  A. Balasuriya,et al.  Autonomous target tracking by AUVs using dynamic vision , 2000, Proceedings of the 2000 International Symposium on Underwater Technology (Cat. No.00EX418).

[5]  S. Gentili,et al.  A hierarchical classification system for object recognition in underwater environments , 2002 .

[6]  G. Conte,et al.  Perception of an underwater structure for inspection and guidance purpose , 1996, Proceedings of the First Euromicro Workshop on Advanced Mobile Robots (EUROBOT '96).

[7]  Gian Luca Foresti,et al.  A Vision Based System for Object Detection in Underwater Images , 2000, Int. J. Pattern Recognit. Artif. Intell..

[8]  Ferial El-Hawary,et al.  Robust regression-based EKF for tracking underwater targets , 1995 .

[9]  Yvan Petillot,et al.  AUTOTRACKER: AUV embedded control architecture for autonomous pipeline and cable tracking , 2003, Oceans 2003. Celebrating the Past ... Teaming Toward the Future (IEEE Cat. No.03CH37492).

[10]  Lotfi A. Zadeh,et al.  From Computing with Numbers to Computing with Words - from Manipulation of Measurements to Manipulation of Perceptions , 2005, Logic, Thought and Action.

[11]  Edward Tunstel,et al.  Enhancing fuzzy robot navigation systems by mimicking human visual perception of natural terrain traversability , 2001, Proceedings Joint 9th IFSA World Congress and 20th NAFIPS International Conference (Cat. No. 01TH8569).

[12]  Tamaki Ura,et al.  Vision-based underwater cable detection and following using AUVs , 2002, OCEANS '02 MTS/IEEE.