English  |  正體中文  |  简体中文  |  Items with full text/Total items : 78852/78852 (100%)
Visitors : 36269979      Online Users : 236
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version

    Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/967

    Title: 影像分割技術於高解析衛星影像分類之應用;Segmentation Technique for Classification of High Spatial Resolution Satellite Images
    Authors: 莊政斌;Cheng-Bing Chuang
    Contributors: 土木工程研究所
    Keywords: 知識庫;影像分割;Image Segmentation;Knowledge-base System
    Date: 2004-06-17
    Issue Date: 2009-09-18 17:16:45 (UTC+8)
    Publisher: 國立中央大學圖書館
    Abstract: 隨著空間解析度的提升,衛星影像的地面特徵更為清晰可辨,但其地物特性亦趨於複雜,相較於著重單一像素光譜值之方法,使用影像分割技術對產生之區塊進行影像的判釋更為合適。本研究之流程分成影像分割以及區塊分類兩個步驟,在影像分割步驟中,使用密度式分割法,利用影像的光譜特性以及空間關聯性進行分割程序,原始影像經影像分割步驟後產生由眾多區塊組成的區塊影像。在區塊分類步驟中,利用知識庫系統,對區塊影像中的各個區塊進行類別的判釋。本研究設計兩組知識庫,一組僅以光譜特性來建立,另一組使用光譜以及紋理兩種特性來建立,分析加入紋理資訊後,其對於影像分類的助益。在影像分割步驟中,使用密度式分割法能簡化影像分割後之區塊數量,降低區塊影像之複雜度,並能提升後續的處理效率,在區塊分類步驟中,利用知識庫系統,能依據影像特性設計多種知識一起進行分析,更有利於影像的分類。本研究使用IKONOS衛星影像作為測試資料,以相同地區之航照判釋土地利用圖作為分類精度評估之依據。成果顯示,僅使用光譜知識庫進行分類之整體精度約為82%,加入光譜資訊後之分類精度可提升為87%,說明紋理資訊對於影像分類具有實質的助益。在高解析衛星影像分類中,使用影像分割方法進行分析是合適的。 As the spatial resolution increases, satellite images are able to provide better land-cover discrimination. However, the content of high resolution images becomes more complicated, pixel-by-pixel image analysis method is not applicable. Therefore, we use image segmentation technique in image process. In this study, the image classification process consists of two steps: image segmentation and object classification. In the first step, algorithm of density-base segmentation is applied, which takes account of spatial connection and the similarity of spectrum space. The process will group pixels into numerous regions and transform the source image into an object image. The second step is designed to classify every region. The classification is based on knowledge-base system which was designed by spectrum and texture information. In this study, we design different knowledge-base system. One of the knowledge-base system consists of spectrum characteristics while the other consists of both spectrum and texture characteristics. The test data of this study are IKONOS satellite images, and the ground truth obtained from aerial photographs is utilized as references for accuracy assessment. The classification accuracy when we use the knowledge-base system consists of only spectrum characteristics is approximately 82%. The classification accuracy e when the knowledge-base system takes texture information into consideration is approximately 87%. For this study, high resolution satellite image classification using segmentation technique and knowledge-base system is applicable.
    Appears in Collections:[土木工程研究所] 博碩士論文

    Files in This Item:

    File SizeFormat

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

    社群 sharing

    ::: Copyright National Central University. | 國立中央大學圖書館版權所有 | 收藏本站 | 設為首頁 | 最佳瀏覽畫面: 1024*768 | 建站日期:8-24-2009 :::
    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - 隱私權政策聲明