Research on Vision-Based Underwater Robot Positioning and Map Building Technology
P. J. Xie †, J. Haleegoah ‡
† Ningbo Dahongying University, Ningbo, 315175, China
‡ CSIR-Science and Technology Policy Research Institute (STEPRI), Accra, Ghana
ABSTRACT: To realize the self-control of the intelligent underwater robot, first of all is to have an independent visual system and the positioning system. Through the visual system, the robot can obtain the underwater environment information, providing the guidance for its movement and underwater works. The vision system of the intelligent underwater robot relies mainly on the “acoustic vision”. The acoustic vision system not only has the acquisition ability of acoustic image and acoustic information, but also processes functions include image and information processing, feature extraction, classification and identification. Meanwhile, the positioning and map building are research hotspot in robot technology field, which is also the key to achieve truly autonomous robot. This paper uses the Super Seaking DST forward-looking sonar to scan the underwater environment and get the sonar image which is needed by the simulation program, and applies the digital image processing method into the sonar image, having processes such as filtering, smoothing, segmentation and so on to the sonar image, extracting the target characteristics and linear characteristics, getting the characteristics map of the underwater environment, constructing the characteristics map simulation platform based on the environmental characteristics, realizing the autonomous positioning and navigation simulation of the AUV by using EKF SLAM algorithm, and having an analysis on AUV movement tracks under the different environmental characteristics and on causes for errors.
Keywords : Underwater robot; Visual system; Forward-looking sonar; Feature extraction; SLAM.
 R. David, R. Pere, “SLAM using an Imaging Sonar for Partially Structured Underwater Environments”, Proc.of IEEE/RSJ International Conference on Intelligent Robots and Systems, 2006, PP. 799-805. Research on vision-based underwater robot positioning and map building technology 216
 W. J. Wang, “EKF SLAM Algorithm in the Application of underwater Vehicle Positioning Research”, Harbin: Institute of Automation of Harbin Engineering University, 2007.
 M. H. Li, B. R. Hong, “Progress of Probabilistic Localization Methods in Mobile Robots”, Robot, vol. 27, no. 4, pp. 380-384, April, 2005.
 Q. Q. Ruan, “Digital Image Processing (second edition)”, Electronic Industry Publishing House, vol. 11, no. 5, pp. 334-336, May, 2007.
 L. X. Yuan, “Underwater Sound Intelligent Robot Visual Imaging Visualization Technology Research”, Harbin: Underwater Acoustic Engineeringof Harbin Engineering University, 2002.
 B. Yan, “Synchronously Positioning and Map Creation of Mobile Robot”, Qingdao: OceanUniversity of China, 2006.
 B. L. Liu, “Simulation System Design for Sonar-based Mobile Robot Mapping”, Journal of System Simulation, vol. 19, no. 13, pp. 43-47, July, 2007.
 C. C. Liu, “High Resolution Imaging Sonar Image Processing. Harbin: Harbin Engineering University”, Signal and Information processing, vol. 25, no. 8, pp. 21-29, August, 2006.  X. L. Hu, K. Y. Xu, “Digital Image Processing”, Beijing: Electronic Industry Press, 2006.