English  |  正體中文  |  简体中文  |  Items with full text/Total items : 28607/40644
Visitors : 5255419      Online Users : 501
RC Version 4.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Adv. Search
LoginUploadHelpAboutAdminister

Please use this identifier to cite or link to this item: http://ntour.ntou.edu.tw:8080/ir/handle/987654321/16222

Title: 非線性濾波器於單相機之運動與形態
Nonlinear filters for Single Camera Based Motion and Shape Estimation
Authors: Jen-Chu Liu
劉仁竹
Contributors: NTOU:Department of Marine Engineering
國立臺灣海洋大學:輪機工程系
Keywords: 剛體運動動態;形態動態;光流動態;卡爾曼濾波器
rigid body motion dynamics;shape dynamics;optical flow dynamics;Kalman filter
Date: 2009
Issue Date: 2011-06-30T08:34:21Z
Abstract: 本文藉由物體在視訊影像上之特徵點,推算出物體移動時的運動與形態,由於串流視訊中,物體在視訊影像上無論是否靜止,有不可避免之雜訊,雜訊的來源有許多種、內部電子元件、外部環境影響等等,因此使用卡爾曼濾波器(Kalman filter)來降低串流視訊上帶來的雜訊所造成之誤差,更精確去測得物體上的特徵點。 在串流視訊裡,物體在現實中所投射在視訊影像上,如何去估測物體移動時的運動與形態,本文利用剛體運動動態(rigid body motion dynamics)以及形態動態(shape dynamics)描述一個非平面之曲面,再配合光流動態(optical flow dynamics)以達到在視訊影像上所呈現的位置,在此系統模型為非線性,故利用擴展型卡爾曼(Extended Kalman filter)和無跡卡爾曼濾波器(Unscented Kalman Filter)對非線性系統做演算,以模擬方式來驗證兩法之可行性,並比較兩法之結果和差異性。 由於擴展型卡爾曼在非線性系統中,其線性化方法必須捨棄高階導數項,導致不穩定狀況,而無跡卡爾曼是由採樣方法對非線性系統進行非線性轉換,不必作線性化,因此可避免線性化之誤差,從結果顯示無跡卡爾曼對於物體在視訊影像上,較能降低誤差且精確測得物體上的特徵點。
The accurate of estimating motion and shape of a moving object is a challenging task because of great variety of noise such as electronic component and influence of external environment etc. In order to alleviate the noise problem, this paper use kalman filter to reduces noise in the streaming video to get more accurate estimation in feature point on the object. Extended Kalman filter have to neglect higher-order derivatives in the nonlinear system to get linearization. However, it will cause unsteady condition. The Unscented Kalman filter uses a deterministic sampling approach to capture the mean and covariance estimates with a minimal set of sample points. UKF can be accurate to the second order for any nonlinearity, avoiding Jacobian’s computation. This paper uses a deterministic sampling of UKF with rigid body motion dynamics, shape dynamics, optical flow dynamics to estimate feature points on the moving object. Results obtained shows that UKF is accurate and reliable to get motion and shape of the object.
URI: http://ethesys.lib.ntou.edu.tw/cdrfb3/record/#G0M96660007
http://ntour.ntou.edu.tw/ir/handle/987654321/16222
Appears in Collections:[輪機工程學系] 博碩士論文

Files in This Item:

File Description SizeFormat
index.html0KbHTML180View/Open


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

 


著作權政策宣告: 本網站之內容為國立臺灣海洋大學所收錄之機構典藏,無償提供學術研究與公眾教育等公益性使用,請合理使用本網站之內容,以尊重著作權人之權益。
網站維護: 海大圖資處 圖書系統組
DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback