Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/64775
Title: 3D reconstruction model
Authors: Peng, Yuhe
Keywords: DRNTU::Engineering
Issue Date: 2014
Abstract: The main component of this project is to realize the 3D reconstruction model of indoor environment. The 3D reconstruction model is an important application of the machine learning. Through the 3D reconstruction model, the unknown environment can be well explored. This is particularly important for human being to execute the task in the dangerous situation. Al so, people can apply this 3D reconstruction model technique into the robotic area so that the robot can execute that kind of tasks instead of human being. In this project research, the method of stereo vision system is proposed. We make use of one camera to take two images from two points of view. Through these two images, we set up the stereo vision system. By analyzing the images and implementing all kinds of image processing algorithm s, we can get the range data and finally realize the purpose of 3D reconstruction model. The platform of this project is the MA TLAB Software. We mainly make use of the Image Processing and the Computer Vision System toolbox in the MATLAB to implement the correspondence feature points searching algorithm s. In the thesis, the geometry of the projection and the triangulation method are introduced. And then, we make an analysis about those kinds of implemented algorithms. The principles of the algorithms are discussed in detail s. The comparison of the algorithm is also give n based on the accuracy and the speed of search. The core component in this project is to map the depth data with the 2 dimensions image data and one kind of efficient mapping method is adopted in this thesis.
URI: http://hdl.handle.net/10356/64775
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
PENG_YUHE_2014.pdf
  Restricted Access
Main report6.89 MBAdobe PDFView/Open

Page view(s) 20

223
checked on Oct 21, 2020

Download(s) 20

15
checked on Oct 21, 2020

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.