淡江大學機構典藏:Item 987654321/88146
English  |  正體中文  |  简体中文  |  全文笔数/总笔数 : 62828/95882 (66%)
造访人次 : 4030468      在线人数 : 830
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
搜寻范围 查询小技巧：
  • 您可在西文检索词汇前后加上"双引号"，以获取较精准的检索结果
  • 若欲以作者姓名搜寻，建议至进阶搜寻限定作者字段，可获得较完整数据
    •  进阶搜寻


    jsp.display-item.identifier=請使用永久網址來引用或連結此文件: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/88146


    题名: 基於SURF之實體物件辨識及其應用
    其它题名: SURF-based object recognition and its application
    作者: 邱俊銘;Chiu, Chung-Ming
    贡献者: 淡江大學電機工程學系碩士班
    謝景棠;Hsieh, Ching-Tang
    关键词: 加速強健特徵;Camshift演算法;擴增實境;無標誌;SURF;Camshift;Augmented Reality;Markerless
    日期: 2012
    上传时间: 2013-04-13 12:02:51 (UTC+8)
    摘要: 博物館一直以來都具有重要的地位，收藏有意義之歷史古蹟、文物、作品，隨著時代演變至今，多媒體影音媒體已成為資訊來源的另一起新秀，其在聲光影音部份都為過去平面書籍所無法提供，如須呈現動態3D、影音呈現的內容時，書籍依然受傳統印刷形式限制而無法呈現上述形式的資訊。因此本文以淡江大學海事博物館為例，期以達到多元的影音經驗與資訊。

    本論文以加速強健特徵(speeded-up robust features, SURF)為基礎，使用CamShift (Continuously Adaptive Mean Shift)對SURF 特徵區域做追蹤。SURF 是在尺度空間(scale space)中尋找穩定點，計算其主要方向，用於匹配特徵點的對齊，而後在鄰近區域內的Haar 小波響應當作紋理特徵進而對特徵點加以描述，因此可以克服物件辨識中影像的縮放大小、旋轉變化、光影變幻等干擾，再利用CamShift來追蹤使用者響導覽的物件，進而顯示擴增實境令使用者有不同的視覺感受．
    The museum has important status in all the time, collection of significant historical sites, artifacts, works with the changing times so far,but along with the time evolution, the multimedia has until now become the information to originate together is in addition beautiful, it all was unable in the acousto-optic video and music part for the past plane books to provide, when like had to present the content which dynamic 3D, the video and music presented, the books still received the traditional printing form to limit to are unable to present the above form the information.


    In this thesis, in order to accelerate the robust features (speeded-up robust features, SURF), based on use CamShift (Continuously Adaptive Mean Shift) SURF feature regions to follow them. SURF is to find a stable point in the scale space (scale space), to calculate the main direction of alignment for matching feature points, then the Haar wavelet response in the neighboring area and then the feature point described as the texture features, so you can overcome the interference of object recognition, image zoom size, rotate, changing light and shadow, re-use CamShift to track the user sound guide to the object, and then display the augmented reality users have different visual experience.
    显示于类别:[電機工程學系暨研究所] 學位論文

    文件中的档案:

    档案 大小格式浏览次数
    index.html0KbHTML138检视/开启

    在機構典藏中所有的数据项都受到原著作权保护.

    TAIR相关文章

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