(44.192.112.123) 您好!臺灣時間:2021/03/08 22:58
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:莊雲翰
研究生(外文):Yun-Han Chuang
論文名稱:結合影像區塊及知識庫分類之研究-以IKONOS衛星影像為例
論文名稱(外文):A research on combining Image Segmentation and Knowledge-based Classification using IKONOS images
指導教授:陳繼藩陳繼藩引用關係
指導教授(外文):Chin-Fan Chen
學位類別:碩士
校院名稱:國立中央大學
系所名稱:土木工程研究所
學門:工程學門
學類:土木工程學類
論文種類:學術論文
論文出版年:2002
畢業學年度:90
語文別:中文
論文頁數:94
中文關鍵詞:知識庫影像分割物件導向
外文關鍵詞:knowledgebaseimage segmentationobject-oriented
相關次數:
  • 被引用被引用:32
  • 點閱點閱:274
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
相較於傳統的影像判識技術著重於對像元光譜值的分析,在本研究的知識庫系統則是採用物件導向的概念來設計,較接近人的辨識過程且具有遺傳性的特性,適合用於分析類別較複雜的影像。本研究的整體處理過程如可概略的分成影像分割與知識庫判釋兩部分;在影像分割的過程中可同時加入光譜與形狀的因子加以約制,原始影像的像元經過影像分割的步驟之後形成許多區塊所組成的區塊影像,每一區塊具有其光譜以及形狀上的屬性。產生區塊影像之後,再利用知識庫來進行辨識的動作,知識庫的設計亦可加入形狀、光譜、遺傳性等條件來幫助判釋,針對影像上的每個區塊計算其在每一類別之歸屬值,完成判釋的工作。本研究是採用IKONOS衛星影像為測試資料,以同一地區的航照人工判識地物分類作為檢核的依據,成果顯示約有80%的分類正確率。
The traditional methods for image interpretation and analysis are based on the spectral characteristics of single pixel. A general idea of object -oriented programming is used in the method we proposed here. Compare to the traditional methods, the object-oriented processing of image information is nearer to the human cognitive process. Furthermore, the hierarchy between objects is well suit for analysis of complex scenes. The image interpretation process consists of two steps, “image segmentation” and “knowledge-based interpretation”. The segmentation will group pixels of image into numerous regions, which form the object image. We can bind the regions of the image by using combination of spectral and spatial factors in segment process. Each region has unique spectral and spatial characteristics. After segmentation, further classification can be achieved by calculating shape, spectral and hierarchy memberships of each region that belongs to each class in the knowledgebase. The test data of this study is an IKONOS satellite image. Using manual-interpreted airplane photo at the same area as reference for accuracy assessment. The correct classified accuracy of this study is about 80 percent.
目錄........ I
圖目錄......III
表目錄...... VI
中文摘要.....VII
Abstract....VIII
第一章序論.....1
1-1 研究動機與研究目的...1
1-2 文獻回顧..............2
1-3 章節簡介.............10
第二章研究方法.............11
2-1 影像處理流程與架構.....11
2-2 影像分割...............12
2-3 知識庫系統.............20
第三章測試影像資料與成果...28
3-1 IKONOS 衛星簡介........28
3-2 測試資料...............30
第四章分類成果精度分析.....58
4-1 土地利用真值圖.........58
4-2 其他分類方式之分類成...63
4-3 分類成果之精度分析.....66
4-4 分類成果討論...........73
第五章結論與建議...........75
參考文獻...................76
附錄A 影像分割操作說明.....78
附錄B 知識庫操作說明.......80
附錄C 土地利用分類圖類別合併表......84
黃俊偉,2001,”大範圍地區土地使用分類之研究”,碩士論文,國立中央大學土木工程研究所,中壢羅秋月,2002,”IKONOS 衛星影像正射改正之研究”,碩士論文,國立中央大學土木工程研究所,中壢DEFINIENS AG , 2000. eCoginition user’s guide.http://www.definiens.com ( July,2001)Hofmann, P., “Detecting urban features from IKONOS data using anobject-oriented approach”, In: Remote Sensing & PhotogrammetrySociety (Editor): Proceedings of the First Annual Conference of theRemote Sensing & Photogrammetry Society 12-14, pp.79-91, SeptemberHofmann, P., “Detecting Informal Settlements from IKONOS image datausing Methods of object oriented image analysis — an example fromCapetown (SOUTH AFRICA)”, JÜRGENS, Remote Sensing of Urban Areas,pp.107-117, 2001.Morisette, J. T. and Khorram, S., “ Accuracy Assessment Curves forSatellite-Based change Detection”, Photogrammetric Engineering &Remote Sensing, Vol. 66, No. 7, pp.875-880, July 2000.Pesaresi, M., and Benediktsson, J. A., ”A New Approach for theMorphological Segmentation of High-Resolution Satellite Imagery”,IEEE translations on Geoscience and Remote Sensing,Vol. 39, No. 2, pp. 309-319, Feb. 2001Richards, J. A., ”Remote Sensing Digital Image Analysis”,Springer-Verlag Berlin Heidelberg, 1986.Schowengerdt, R. A., “Remote Sensing models and methods for imageprocessing”, 2nd , Academic Press, 1997.Tonjes, R., Growe, S., Bucker, and Liedtke, C.-E., ”Knowledge-BasedInterpretation of Remote Sensing Images Using Semantic Nets ”,Photogrammetric Engineering & Remote Sensing, Vol. 65, No. 7,pp.811-821, July 1999.Takashi, K., Hiroichi, E., and Yoshiyuki, K., ”Classification of the SPOTImage Using Spectral and Spatial Features of Primitive Regions ThatHave Nearly Uniform Color”, IEEE translations on Geoscience andRemote Sensing, Vol.28, No. 4, pp.749-752, July 1990.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔