English  |  正體中文  |  简体中文  |  全文筆數/總筆數 : 62797/95867 (66%)
造訪人次 : 3749388      線上人數 : 464
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
搜尋範圍 查詢小技巧:
  • 您可在西文檢索詞彙前後加上"雙引號",以獲取較精準的檢索結果
  • 若欲以作者姓名搜尋,建議至進階搜尋限定作者欄位,可獲得較完整資料
  • 進階搜尋
    請使用永久網址來引用或連結此文件: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/103137


    題名: 應用模糊群聚演算法與群聚驗證之消失點偵測研究
    其他題名: Vanishing-Point Detection Based on Fuzzy Clustering Algorithm and New Clustering Validity Measure
    作者: 周建興
    貢獻者: 淡江大學電機工程學系
    日期: 2012-08
    上傳時間: 2015-05-19 15:44:38 (UTC+8)
    摘要: 3D 立體影像在未來的科技生活中,扮演越來越重要的角色,有關3D 影像的開發以 及應用,是目前相當受到重視的研究方向。此計畫主要探討的研究主題,是使用深度 估測法將單張2D 影像轉換成3D 影像。深度估測法最重要的目標,就是從一張影像中 估測出場景的深度圖;然後使用原始影像與深度圖產生出左眼與右眼的影像。在整個 流程中有兩個最重要的步驟,一個是將前景物體切割出來,另一個是指派適當的深度 值給所有的物體。但是,要在所有類型的影像中,成功地切割出前景物體,或是估計 出影像的深度圖,都可說是十分困難的任務。 在此計畫中,我們首先結合模糊群聚演算法,提出一個影像前處理方法。所提出 的前處理方法可以在影像中擷取出較關鍵的邊界資訊,以偵測出消失線與消失點的位 置。此外,在這計畫中,我們也預計提出一個全新的群聚驗證演算法,這個群聚驗證 演算法將評估群聚的面積是否相近,以估測出合適的群聚數目。依據估測到的群聚數 目,搭配所提出的前處理方法,我們能更準確地定位出消失點與消失線的位置,還能 進一步提供影像中前景與背景的資訊,有助於建立正確的深度圖。
    Three-dimensional images will play an increasingly important role in the future of science and technology. The development and application of 3D images, therefore, is currently a major focus of research. This project focuses on converting a single 2D image into a 3D image using the depth-estimation method. The objective of a depth-estimation method is estimating the depth map of a scene from an image; the left-eye and right-eye images can then be produced using the original image and depth map. Two critical steps are included in this process. One segments the foreground objects and background; the other assigns appropriate depth value to all objects. However, it is difficult to design a method to segment foreground objects and generate depth maps for all types of images. In combination with the fuzzy clustering algorithm, this project proposes an image-preprocessing method. Critical edge information can thus be extracted from an image to locate vanishing lines and the vanishing point. To estimate an appropriate initial cluster number, this project also proposes a new clustering validity measure to assess the similarity of cluster areas. By applying the estimated cluster number and the proposed processing method, we can not only accurately discover the location of the vanishing point and vanishing lines, but also further provide useful information for image segmentation.
    顯示於類別:[電機工程學系暨研究所] 研究報告

    文件中的檔案:

    沒有與此文件相關的檔案.

    在機構典藏中所有的資料項目都受到原著作權保護.

    TAIR相關文章

    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - 回饋