碩士 / 國立臺灣海洋大學 / 通訊與導航工程系 / 96 / Fishing Vessel Monitoring System (VMS) is an effective tool of fisheries monitoring, control and surveillance measures to counter over-fishing. It can also help the coast guard to safeguard vessels more efficiently. As VMS is widely implemented, more and more efforts focus on mining the VMS database to discover knowledge and clues that would further enhance the benefits. This thesis is focused on data mining VMS database with clustering technology developed for and implemented into the VMS of Taiwan. The initial request form the Fisheries Administration was to constantly identify wherever there are at least three fishing vessels within 3 nautical miles of range. The proposed solution was based on DBSCAN [1] clustering algorithm. The performances in accuracy and run-time were evaluated and improved with vessel position prediction, partitioning of datasets, data structure and algorithm design. With the promising results, this solution has been recognized by the fisheries management and VMS operation experts to be of many extended use in VMS.
Finally, this Density Area Detection System was applied to the detection of at-sea transshipment and parallel-track vessels. Then, the performance in accuracy and practicability would be discussed.
Identifer | oai:union.ndltd.org:TW/096NTOU5300018 |
Date | January 2008 |
Creators | Ying-Yuan Su, 蘇膺元 |
Contributors | Shwu-Jing Chang, 張淑淨 |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 84 |
Page generated in 0.0014 seconds