City University of Hong Kong

CityU Institutional Repository >
3_CityU Electronic Theses and Dissertations >
ETD - Dept. of Manufacturing Engineering and Engineering Management  >
MEEM - Doctor of Philosophy  >

Please use this identifier to cite or link to this item:

Title: Active vision perception : sensor self-recalibration, automatic sensor placement, and 3D reconstruction
Other Titles: Zhu dong ji xie shi jue : gan zhi qi de zi fu biao ding, zi dong qu wei, ji san wei chong jian
主動機械視覺 : 感知器的自複標定, 自動取位, 及三維重建
Authors: Chen, Shengyong (陳勝勇)
Department: Dept. of Manufacturing Engineering and Engineering Management
Degree: Doctor of Philosophy
Issue Date: 2003
Publisher: City University of Hong Kong
Subjects: Calibration
Computer vision
Image processing
Image reconstruction
Three-dimensional imaging
Visual perception
Notes: CityU Call Number: TA1634.C44 2003
Includes bibliographical references (leaves 215-232)
Thesis (Ph.D.)--City University of Hong Kong, 2003
x, 232 leaves : ill. ; 30 cm.
Type: Thesis
Abstract: An active visual perception system can change its visual parameters in an intentional manner and perform its sensing actions purposefully. A general vision task thus can be performed in an efficient way by means of strategic control of the perception process. The controllable processes include 3D active sensing, sensor configuration and recalibration, automatic sensor placement, and 3D reconstruction. This thesis presents the research in studying the relevant issues for active visual perception. Compared with passive vision sensors, active sensors feature high reliability and accuracy. Structured lighting is one of the widely used active methods and is suitable for industrial applications. This thesis considers two types of setups, the stripe light vision system and color-encoded vision system. Both of them are reconfigurable during the active perception. In order to reconstruct an object with high accuracy, it is essential that the vision sensor be carefully calibrated. Traditional calibration methods are mainly for static uses in which a calibration target with specially designed features needs to be placed at precisely known locations. In practical applications, the configuration of a vision system often needs to be changed to meet the constraints in different views, in which case the vision system must be recalibrated. This is particularly the case for visual perception using an active vision system in which the sensor needs to be moved arbitrarily to observe different object features. To avoid the tedious and laborious procedures in traditional calibration, a self-recalibration method is developed in this thesis. The new method dynamically calibrates the active vision system when and if the relative configuration is changed. A distinct advantage of this method is that neither an accurately designed calibration device nor the prior knowledge of the motion of the camera or the scene is required. Vision sensors have limited fields of views and can only "see" a portion of a scene from a single viewpoint. To make the entire object visible, the sensor has to be moved from one place to another around the object to observe all features of interest. The sensor placement presented in this thesis describes an effective strategy to generate a sequence of viewing poses and sensor settings for optimally completing a perception task. Several methods are proposed to solve the problems in both model-based and non-model-based vision tasks. For model-based applications, the method involves determination of the optimal sensor placements and a shortest path through these viewpoints for automatic generation of a perception plan. A topology of viewpoints is achieved by a genetic algorithm in which a min-max criterion is used for evaluation. A shortest path is also determined by Christofides algorithm. For nonmodel-based applications, the method involves determination of the best next view and sensor settings. The trend surface is proposed as the cue to predict the unknown portion of an object or environment. Experiments are conducted with an active visual sensing system consisting a pattern projector and a CCD camera. The results obtained show the validity of the proposed methods and the feasibility for practical implementation. Using the active visual perception strategy, 3D reconstruction can be achieved without constraints on the system configuration parameters. This allows optimal system configuration to be employed to adaptively sense an environment. Together with the optimal sensor placement strategy, the proposed methods will give the vision system the adaptability needed in many practical applications.
Online Catalog Link:
Appears in Collections:MEEM - Doctor of Philosophy

Files in This Item:

File Description SizeFormat
fulltext.html159 BHTMLView/Open
abstract.html159 BHTMLView/Open

Items in CityU IR are protected by copyright, with all rights reserved, unless otherwise indicated.


Valid XHTML 1.0!
DSpace Software © 2013 CityU Library - Send feedback to Library Systems
Privacy Policy · Copyright · Disclaimer