Real Time Seafloor Classification Using Sidescan Sonar Scanlines for Unmanned Underwater Vehicle Navigation

• Main Authors: Sheng-Wei Huang, 黃盛煒
• Other Authors: 郭振華
• Format: Others
• Language: en_US
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/11109445430044559980

碩士 === 國立臺灣大學 === 工程科學及海洋工程學研究所 === 96 === Unmanned Underwater Vehicles (UUVs) have been employed in tasks such as hydrographic survey, underwater surveillance. UUVs are stable sensor platforms with positioning capabilities. However, navigation of UUV in unstructured environments is a still challenging problem due to the lack of predefined landmarks on the sea floor. Developing real time seafloor features detection capability will be beneficial for the UUV navigation. The aim of this work is the development of a real-time method of target detection by the processing of side-scan sonar sscanlines. The detection process has three parts. Firstly, a pre-processing step is used to filter the data and to reduce computation time. Then a detector is designed to categorize the target content in a scanline and represents targets with an adaptive threshold. Finally, target positions are corrected based on the motion of the UUV. This procedure was implemented on an UUV to verify the landmark detection capability in real time. Two experimental results conducted on sandy seabed with concrete targets are demonstrated to show their signal contents and the final grid maps that are ready for UUV navigation purpose. The detection method developed in this work enables UUVs to build sea floor map for the self localization in an unstructured environment.