Image matching for 3D reconstruction
In this dissertation, we introduce LoGLUE, a novel method of image matching for 3D reconstruction that produces results for matching images with low texture. Compared with current image matching methods, we design a backbone with the Convolutional Neural Network (CNN) to extract both coarse- and fi...
Main Author: | Zhang, Yixuan |
---|---|
Other Authors: | Tan Yap Peng |
Format: | Thesis-Master by Coursework |
Language: | English |
Published: |
Nanyang Technological University
2023
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/172862 |
Similar Items
-
3D stereo reconstruction using sum square of different matching algorithm
by: Marghany, Maged, et al.
Published: (2011) -
3D model reconstruction from 2D images
by: Fong, Yi Qing
Published: (2019) -
3D reconstruction and visualization of medical images : computed tomography
by: Tee, Yvonne Qiying.
Published: (2009) -
Stereo Camera - Based 3D Object Reconstruction Utilizing Semi-Global Matching Algorithm
by: Achmad, M. S. Hendriyawan, et al.
Published: (2016) -
3D reconstruction of ocular fundus images
by: Liu, Yang
Published: (2008)