English  |  正體中文  |  简体中文  |  Items with full text/Total items : 28611/40649
Visitors : 650588      Online Users : 46
RC Version 4.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Adv. Search
LoginUploadHelpAboutAdminister

Please use this identifier to cite or link to this item: http://ntour.ntou.edu.tw:8080/ir/handle/987654321/33657

Title: 應用遙測影像資料於台灣崩塌地變遷及其敏感度分析之研究
Authors: 張陽郎;方志鵬;梁文耀;謝東儒;張麗娜
Contributors: 國立臺灣海洋大學:通訊與導航工程學系
Keywords: 崩塌地變遷敏感度;地質遙測學;粒子群尋優化奇異值分解特徵萃取;以KD 樹為基礎之準擬陣分類器;以凸包為基礎之多類分類器;動態分散式平行計算環境即時嵌入式
landslide change detection;landslide susceptibility analysis;particle swarmoptimization-singular value decomposition;KD-tree based semi-matroid classifier;convex hull based semi-matroid classifier;dynamic distributed mobile computingenvironment
Date: 2009-08
Issue Date: 2013-05-08T06:36:42Z
Publisher: 行政院國家科學委員會
Abstract: 摘要:由於人為大量的污染破壞,導致「溫室效應」(the greenhouse effect) 持續擴大,逐年改變全 世界的氣候,將造成全球生態環境極大的變遷。台灣地處亞熱帶,地理位置與生態環境特殊, 加上地勢陡峭,經常性的颱風豪雨,及過度的人為開發,使得水土保持、森林保育的工作需 長期持續的監測、養護及規劃。台灣環境變化原因錯綜複雜,不斷地影響到臺灣崩塌地的變 遷,如何運用有系統的「災害管理」及「地質遙測」技術,來有效管理崩塌地的變遷,將是 我們急需克服的課題。尤其是近年來重大災害頻傳,九二一震災後,崩塌裸露區激增,接踵 而至的颱風帶來連續的激增雨量,更就造成了新的崩塌與災害,例如2004 年的敏督利颱風, 其超強的雨量,造成了九二一地震後鬆動的崩塌地,更為嚴重的土石流災害。因此,本三年 期研究計畫提出一個利用遙測影像分析「崩塌地敏感度分類」的新方法,用以研究台灣地區 「崩塌地變遷」及「災害管理」的問題,並希望藉由先前研究所累積之寶貴經驗,尋找出一 個適合應用於「崩塌地敏感度」的最佳分類方法,發展出另一套新的最佳平行「粒子群尋優 化奇異值分解」(Particle Swarm Optimization - Singular Value Decomposition, PSO-SVD) 的 「崩塌地敏感」特徵萃取演算法、「以KD 樹為基礎之準擬陣分類器」(KD-tree based Semi-Matroid Classifier) 及「凸包準擬陣分類器」(Convex Hull-based Semi-Matroid Classifier)」 演算法,用以增進特徵萃取的最佳化效益,進而提升遙測影像「崩塌敏感度」分類辨識及變 遷預測率,最後再整合以上方法,實作於「動態分散式平行計算環境即時嵌入式」系統中。 希望藉由嚴謹的理論探討及深入的推導分析,以實作的結果,驗證所提之新架構優於傳統的 地質遙測分類及變遷預測方法。尤其近年來台灣飽受氣候異常,地震、颱風等天然災害的肆 虐,造成部分地質及土壤嚴重的土石流現象,不但危害居民的生命財產,且破壞了環境的自 然生態,如何利用最新的遙測影像分類處理方法,有效控制天然災害及管理土地資源,並藉 由整合「地球科學」、「地質科學」、「遙測科技」、「資訊科學」及「電機工程」等不同學科研 究,來跨領域探討台灣地區特有的「崩塌地變遷」與「災害管理」等相關的議題。
abstract:A novel study is proposed for automatic landslide change detection and landslide susceptibility analysis on Taiwan using geological remote sensing images. The planning method is based on fusion of high-dimensional remote sensing images of the same scene collected from multiple sources. It will present a framework for fusion of multisource remote sensing images, which consists of three algorithms, referred to as particle swarm optimization - singular value decomposition (PSO-SVD) for band selections, KD-tree based semi-matroid classifier (KDSM) and convex hull based semi-matroid classifier (CHSM). Based on our previous proposed complete modular eigenspace (CME) method which was designed to extract the simplest feature modules, the PSO-SVD is intended to improve the performance of the extracted CME features optimally by modifying the original CME using PSO algorithm and further decomposing them into the most efficient feature subspaces before applying to classification phases. The KDSM and CHSM classifiers are designed to fit the selected PSO-SVD features and further optimize the landslide classification accuracies. The performances of these proposed methods, implemented by a new defined dynamic distributed mobile computing environment (D2MCE) in a real-time embedded system, is evaluated by fusing MODIS/ASTER (MASTER) hyperspectral, synthetic aperture radar (SAR) and light detection and ranging (LiDAR) images. It’ll be expected that the experiments will demonstrate the availability of the proposed PSO-SVD/KDSM/CHSM approaches. PSO-SVD is an effective method for the feature extractions of the high-dimensional fused data. KDSM/CHSM can also improve the classification accuracies and prediction rates of landslide susceptibility significantly compared to the conventional geological remote sensing landslide change detection and landslide susceptibility analysis methods.
Relation: NSC98-2116-M027-002
URI: http://ntour.ntou.edu.tw/handle/987654321/33657
Appears in Collections:[通訊與導航工程學系] 研究計畫

Files in This Item:

File Description SizeFormat
index.html0KbHTML173View/Open


All items in NTOUR are protected by copyright, with all rights reserved.

 


著作權政策宣告: 本網站之內容為國立臺灣海洋大學所收錄之機構典藏,無償提供學術研究與公眾教育等公益性使用,請合理使用本網站之內容,以尊重著作權人之權益。
網站維護: 海大圖資處 圖書系統組
DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback