English  |  正體中文  |  简体中文  |  Items with full text/Total items : 78852/78852 (100%)
Visitors : 37591923      Online Users : 3220
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/987654321/9490

    Title: 特徵顏色表示方法及其在物體偵測上之應用;Eigen Color Representation and Its Applications to Object Detection
    Authors: 蔡洛緯;Luo-Wei Tsai
    Contributors: 資訊工程研究所
    Keywords: 特徵顏色;物體偵測;traffic sign detection;vehicle detection;eigen color;object detection
    Date: 2009-01-09
    Issue Date: 2009-09-22 11:48:55 (UTC+8)
    Publisher: 國立中央大學圖書館
    Abstract: 在電腦視覺領域中,物體偵測是相當基礎且重要的問題。同時可應用在很多方面,例如:視訊監控,導航,影像檢索…等。主要目的是找出物體在影像中的正確位置不論場景如何地變化. 本論文提出一套新穎的系統架構應用於彩色影像中。首先, 我們發展出一種稱做特徵值顏色的方法。此方法是透過對某特定物體類別做一統計上的分析所推導得到的結果.在這個新的特徵色彩空間上,前景物像素點可以容易地與背景物的像素點作區分,即使是在一些具有光線變化的場景。至於在候選區塊的確認步驟,我們利用數種重要的物體外觀特徵包含角點、邊緣資訊與小波轉換之係數,來建構一串連且多重維度之物體分類器。依據此串連架構,可以對輸入影像中可能的前景物像素點作有效之確認。由於先前已利用色彩資訊濾除大量無關的背景像素點,故此掃瞄步驟將可快速的執行並找出前景物。 與一般傳統外觀類型的偵測方式相比,我們所提出的特徵色彩空間可以事先過濾大量無關的背景像素點.因此可以有效的快速定位出物體的位置。即使是靜態影像,我們仍舊可以成功的從非固定式的照相機偵測出前景物。我們分別利用車輛與交通號誌的偵測來驗證所提出方法的可行性。實驗結果證明結合特徵色彩資訊與局部外觀資訊之偵測方式是強而有效的。 Object detection is a fundamental and important problem in computer vision and can be applied to various applications like video surveillance, navigation, content-based image retrieval and so on. Its goal is to find the exact location of an object no matter how the environmental conditions change. This thesis presents a novel framework for detecting objects in color images. First of all, a novel eigen color representation derived from a statistical analysis of object instances is presented. In this new eigen-color space, different object pixels can be easily identified from background, even though they are lighted under varying illuminations. At the hypothesis verification stage, each detected pixel corresponds to an object hypothesis. Several important appearance features including corners, edge maps and coefficients of wavelet transforms were used for constructing a cascade multi-channel classifier. With the cascade structure, an effective scanning process can be performed to verify all possible candidates. Because the color feature eliminates most background pixels in advance, the scanning process can be performed extremely quickly to locate each desired object. Compared with the traditional appearance-based methods, our proposed eigen-color space can filter out most of impossible candidates in advance and thus each desired object can be very efficiently located from the background. Even thought still images are handled, each object still can be efficiently detected from a non-stationary camera. Two important applications are demonstrated in this thesis; that is, vehicle detection and road sign detection. Experimental results demonstrate that the integration of eigen color feature and local appearance features can form a powerful and superior tool in object detection.
    Appears in Collections:[資訊工程研究所] 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