3D reconstruction using image features

For more info about our services contact : help@bestpfe.com

Table of contents

1 Introduction 
1.1 Multi-view stereo for 3D reconstruction
1.1.1 Classication of Image-based 3D reconstruction methods
1.2 Main contributions and thesis outline
1.2.1 Main contributions
1.2.2 Thesis outline
2 Image matching for 3D reconstruction 
2.1 Introduction
2.2 Local methods
2.3 Global methods
2.3.1 Variational techniques for global methods
2.3.2 Discrete techniques for global methods
2.3.3 Evaluation data-sets
3 Feature detection and matching 
3.1 What are local features?
3.2 Feature detection
3.2.1 Curvature based feature detectors
3.2.2 Intensity based feature detectors
3.2.3 Segmentation based features
3.2.4 Model based feature detectors
3.2.5 Color based feature detectors
3.3 Feature descriptors and matching
3.3.1 Feature descriptors based on distribution
3.3.2 Feature descriptors based on lters
3.3.3 LDAHash
3.3.4 Feature descriptors based on spin images
3.3.5 Feature descriptors based on color
3.4 3D reconstruction using image features
3.4.1 DAISY
3.4.2 SIFT Flow
3.5 Experiments
3.6 Conclusion
4 Variational Image Matching for 3D reconstruction 
4.1 Classical formulation
4.2 Data term
4.2.1 Data term penalizers
4.2.2 What to measure
4.3 Regularization
4.4 Data term linearization
4.4.1 Early data term linearization
4.4.2 Late data term linearization
4.5 Additional cues from scene geometry
4.5.1 Epipolar geometry
4.5.2 Sparse feature matches
4.6 Solving the variational matching problem
4.6.1 Successive over relaxation: SOR
4.6.2 Alternating direction implicit: ADI
4.6.3 Coarse to ne
4.6.4 Multi-grid
4.7 Experiments
4.7.1 Datasets
4.7.2 Methods and criteria
5 Distortion driven variational multi-view reconstruction 
5.1 Introduction
5.2 Geometry driven variational matching
5.2.1 Classical variational matching
5.2.2 Distortion driven variational matching
5.2.3 Results using distortion driven variational matching
5.3 Distortion driven multiple-view merging
5.3.1 Triplet contributions
5.3.2 Pair contribution
5.4 Results and discussion
5.5 Conclusion
6 Propagation-based matching for 3D reconstruction 
6.1 Introduction
6.2 Match propagation for wide-baseline congurations
6.3 Quasi-dense match propagation for wide-baseline congurations
6.3.1 Quasi-Dense Wide Baseline Matching for Three-Views
6.3.2 Multi-view quasi-dense matching for wide-baseline congurations
6.3.3 Results and discussion
6.4 Accurate, Dense, and Robust Multiview Stereopsis
6.4.1 Results and discussion
7 Complementary geometric and optical information for match prop- agation based 3D reconstruction 
7.1 Introduction
7.2 Geometry based image match propagation
7.2.1 Overview
7.2.2 Propagation
7.2.3 Candidate region querying
7.2.4 Fit surface and update
7.3 Results
7.4 Conclusion and discussion
8 Conclusion and perspective 
8.1 Conclusion
8.2 Perspective
Bibliography