查詢結果分析
相關文獻
- An Object-Based Method to Integrate Remote Sensing and Geographical Information for Advanced Feature Extraction
- 從地理期刊之發表探討臺灣近卅年來地圖學之發展趨勢
- 整合地理資訊系統與遙測技術於林分材積估側之研究
- 利用地理資訊系統和衛星遙測估算林分材積
- 遙測與地理資訊系統應用於集水區水文模擬之技術探討
- 整合遙測、地理資訊系統及衛星定位系統在自然資源經營的應用--美國科羅拉多州立大學森林系Roger M. Hoffer教授演講摘要
- State-of-the-art Technologies of Forest Inventory and Monitoring in Taiwan
- 衛星遙測與GIS於縣市政府山坡地監測管理
- 賀伯颱風災害後嘉義山坡地新增裸地之空間分析
- 遙測影像(RS)、地理資訊系統(GIS)、全球衛星定位儀(GPS)應用在森林火災之研究
頁籤選單縮合
題名 | An Object-Based Method to Integrate Remote Sensing and Geographical Information for Advanced Feature Extraction=以物件為基礎的方法整合地理資訊及遙測影像做土地使用特徵擷取與分類 |
---|---|
作者 | 王聖銘; Wang, Sheng-ming; |
期刊 | 臺灣地理資訊學刊 |
出版日期 | 20050400 |
卷期 | 2 民94.04 |
頁次 | 頁11-26 |
分類號 | 440.98 |
語文 | eng |
關鍵詞 | 地理資訊系統; 遙測; 影像分析; 特徵擷取; 土地使用分類; Geographical information system; Remote sensing; Image analysis; Feature extraction; Land use classification; |
中文摘要 | 本文在呈現利用以物件為基礎的方法整合地理資訊及遙測影像資料做土地使用特徵擷取與分類的研究。本研究的動機在於探討如何整合多樣性的空間資料,並轉化為物件化的知識,作為土地使用特徵擷取與分類的基礎。在本研究中所提出之以物件為基礎分析單元並結合物件內含知識推論的分析方法,比傳統的以像元或網路區域為基礎分析單元的分析方法,更接近真實世界認知與分類方法,並可獲得較好的資訊擷取與分類結果。 本研究依相關分析與設計,建構了實際應用的雛型系統-OMIRGS。並利用兩組不同年份的空間資料做案例分析。由案例分析的結果顯示,利用OMIRGS針對本研究中所設定的九種土地使用特徵擷取及分類的結果,分別有93%及94%的分類準確度。相對於傳統的單純以SMAP關聯性影像分類方法所的到的結果,其結果有顯著的改進。而由相關結果顯示,利用本研究所設計,以物件為基礎分析單元並結合物件內含知識推論的分析方法,能擷取到僅利用傳統的遙測影像分類分析方法所無法得到的分類結果。 |
英文摘要 | This paper presents an object-based method to integrate remote sensing and geographical information systems for advanced feature extraction. The research was motivated by the need to integrate spatial data for feature extraction and the hypothesis that an object-based approach would lead to improved results compared to pixel and kernel-based approaches. A prototype system –OMIRGS, had been designed and implemented in this research. A case study for object land use classification and change detection within an urban-rural area has been carried out. Promising results have been obtained from the case study. The 1988 and 1990 land use maps of the case study area, with 9 classes of land use features; have been generated with an accuracy of 93% and 94%. Significant improvements have been obtained comparing the results with those generated using the SMAP contextual image classification method alone. This research concludes that the method proposed in this research can be used to extract the features, which are very difficult to derive using conventional techniques and remotely sensed data alone. The work contributes to the prioritized research area of the integration of remote sensing and geographic information systems for advanced feature extraction. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。