Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/84769
標題: 建立流域影像分類之規則集與其應用之可行性評估
Establishment of the Watershed Image Classified Rule-Set and Feasibility Assessment of Its Application
作者: 林承翰
莊心凱
林銘郎
Cheng-Han Lin
Hsin-Kai Chuang
Ming-Lang Lin
關鍵字: 流域衛星影像ll物件式影像分析ll規則集ll地工災害;Watershed satellite imagellobject-based image analysisllrule-setllgeotechnical damage
出版社: 台北市:中華水土保持學會
Project: 中華水土保持學報, Volume 44, Issue 3, Page(s) 215-224.
摘要: 
Extreme weather disasters are widely distributed. Moreover, the mutual relations between such disasters can easily make them even more extensive. Rapidly obtained satellite images of vast areas can be effective in helping to develop disaster prevention strategies. This study established a watershed satellite image classification rule-set by using the object-based image analysis methodology. Several surface features in the watershed image were been classified, including main channels, secondary channels, sandbars, alluvial fans, landslides and place of the geotechnical damage.
This study practically applied this rule-set in different watersheds and different-resolution satellite image. Also assessed the feasibility of the rule-set by comparing with the investigation photos and statistical analysis. The results show that the rule-set logic can be applied flexibly in different watersheds and different images. The classification of the rule-set is reproducible and accurate. As this result, we can apply the rule-set to disaster management and the land use planning in the future work.

極端氣候下的災害事件分布廣泛,且相互影響之關聯性易使災害規模擴大,藉由衛星
影像判釋具廣域且快速的特色能夠有效協助研擬防災策略。本研究建立一套流域衛星影像分類
之規則集,當中引入物件式影像分析方法,使分類流程依據人工判釋之經驗與知識,針對流域
影像分類出周邊河川地形單元、崩塌地以及地工災害的位置。實際應用此規則集於不同流域以
及不同解析度之衛星影像,透過比對現地勘查資料和計算誤差矩陣來評估其邏輯之可行程度。
結果顯示,本研究所建立之規則集邏輯泛用於不同流域、不同影像,且任何使用者之操作經過
適當的調整皆具有重現性及相當的判釋精度,提供災害管理、土地利用之深入探討一項工具。
URI: http://hdl.handle.net/11455/84769
Appears in Collections:第44卷 第03期

Files in This Item:
File Description SizeFormat
84614-3.pdf18.65 MBAdobe PDFView/Open
Show full item record
 

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.