English  |  正體中文  |  简体中文  |  Items with full text/Total items : 55025/89277 (62%)
Visitors : 10605406      Online Users : 19
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU 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://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/103117

    Title: 視覺式即時建圖與定位之改進與硬體實現
    Other Titles: Visual Slam Improvement and Implement
    Authors: 孫崇訓;王銀添
    Contributors: 淡江大學機械與機電工程學系
    Keywords: 機器視覺;加速強健特徵(SURF);同時定位與建圖(SLAM);擴張型卡爾曼過濾器(EKF);現場可編程輯閘陣列(FPGA);machine vision;Speeded-Up Robust Features (SURF);Simultaneous Localization And Mapping (SLAM);Extended Kalman Filter (EKF);Field-Programmable Gate Arrays (FPGA)
    Date: 2012-08
    Issue Date: 2015-05-19 13:33:49 (UTC+8)
    Abstract: 若要讓一個移動機器人能達到真正的自主移動,在其移動環境中的地圖建置與定位能力可說是非常重要的議題。本計畫將基於加速強健特徵(SURF)與擴張型卡爾曼濾波器(EKF)的視覺式同時定位與建圖(SLAM)以現場可編程輯閘陣列(FPGA)實現。移動機器人的控制也將由FPGA系統完成。我們簡化了SURF的描述向量與興趣點搜尋以及採用階層式SLAM,以減少FPGA系統的運算負擔。在此計畫中,我們採用雙眼視覺實現視覺式SLAM。以左眼攝影機擷取到的影像進行狀態的預測與量測,新增地標時則使用左右攝影機所構成的立體視覺求算影像深度,以加速新增特徵的初始化。在SLAM的資料關聯與地圖管理策略,我們採用移動視窗的概念,以提高比對的效率。最後我們將這些基於效率考量步驟整合在FPGA系統上,以實現移動機器人的自主移動。
    Localization and mapping are two of the most important issues to make a mobile robot truly autonomous. In this project, a visual simultaneous localization and mapping (SLAM) is established by the extended Kalman filter (EKF) and speeded-up robust features (SURF). The EKF-based visual SLAM will be developed and implemented on FPGA-based architecture. Also, control for the mobile robot will be realized by the FPGA-based architecture. In order to reduce the computation load to the FPGA architecture, the descriptor and detection of interest points of the SURF algorithm are modified. The concept of hierarchical SLAM is considered to reduce the computation demand. The EKF SLAM is implemented on a binocular vision system. The state measurement and estimation are according to the left camera captured images. The image depth of a new landmark is derived under the standard stereo geometry architecture which is composed of the captured images by left and right cameras. This binocular vision structure also takes low computation demand. An efficient searching window is proposed to accelerate the procedures of data association and map management for EKF SLAM. Finally, the aforementioned procedures are integrated and implemented on FPGA-based architecture and the autonomous mobile robot.
    Appears in Collections:[Graduate Institute & Department of Mechanical and Electro-Mechanical Engineering] Research Paper

    Files in This Item:

    There are no files associated with this item.

    All items in 機構典藏 are protected by copyright, with all rights reserved.

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