應用支援向量機於九份二山崩塌地變遷評估之研究

Using Data Mining Techniques for Landslide Change Assessment at the Chiufanershan Area

林文賜(Wen-Tzu Lin) ； 廖紹安(Shao-An Liao) ； 張文詔(Wen-Chao Chang) ； 周建明(Chien-Ming Chou) ； 洪耀明(Yao-Ming Hong)

《水保技術》 3卷3期 (2008/06) Pp. 108-116

https://doi.org/10.30040/JSWCT.200806.0001

摘要

921集集大地震造成中部地區大規模崩塌，地震引起之崩塌區位分遼闊且零散，本研究以南投縣國姓鄉九份二山爲樣區，以921地震後之SPOT衛星影像分析崩塌地之燮遷資訊。首先藉由影像相減法分析崩塌地之光譜特性，並輔以資料探勘技術之支援向量機分析不同時期之崩塌區位及分類準確度，作爲崩塌地燮遷分析及災害治理之參考。九份二山崩塌面積由地震初期(1999/9/27)之195.7公頃，於九年後(2007/5/7)減少爲47%公頃，約有75.5%崩塌區位復原，在經九年後之時間，植生陵績入侵崩塌地，崩塌區位有逐年復原之趨勢。

關鍵字

崩塌地 ； 資料探勘技術 ； 支援向量機

並列摘要

Massive landslides, caused by the catastrophic Chi-Chi earthquake on September 21, 1999, at Chiufanershan area in Nantou County. In this study, multi-temporal SPOT satellite images were chosen for landslide change analysis. Firstly, image subtraction method was employed for analyzing landslide spectral characteristics. Secondly, the landslide sites at different dates were identified using support vector machine (SVM). The analyzed results indicate that landslide areas have decreased from 195.7 ha on September 27, 1999 to 47.96 ha on May 5, 2007, about 75.5% of area restored, indicating that the sites of landslide have been gradually restored over nine years of natural vegetation.

並列關鍵字

Landslide ； Data mining technique ； Support vector machine