de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Differt, Dario: Holistic methods for visual navigation of mobile robots in outdoor environments. 2017
Inhalt
Abstract
Acknowledgements
Table of Symbols
Introduction
Autonomous Navigation
Visual Navigation
Visual Localization
Visual Homing
Route Following
Visual Compass
Biologically Inspired Visual Navigation
Omnidirectional Camera Sensors
Outline
Multispectral Skyline Extraction
Introduction
Navigational Abilities of Social Insects
Skyline as Landmark Cue
Perception of Light
Global and Local Classification Methods
Related Work
Contributions
Materials and Methods
Experimental Setup
Calibration
Data Collection
HDR Imaging
Creation of Data Samples
Data Visualization
Classification Rate
Data Classification
Overview: Tested Separation Techniques
Results
Collected Data
Global Separation Techniques
Local Separation Techniques
Comparison between Global and Local Separation Techniques
Statistical Tests
Records of Ground Objects
Panoramic Images
Discussion
Skyline Extraction
Panoramic Images
Color Contrast Mechanisms in Insects
Future Work
Conclusion
Spherical Harmonics: Theory & Software Implementation
Introduction
Motivation
Rotation Group SO(3)
Elementary Rotation Matrices
Tilt Matrices
Distance Measure for Rotation Matrices
Fourier-Transform and Spectra
Fourier Analysis on SO(3)
Wigner-D matrices
Clebsch-Gordan Matrices
Spherical Harmonics
Alternative Formulation of Spherical Harmonics
Point-Wise Products
Real Spherical Harmonics
Recurrence Relations
Symmetries
Rotations
Translations
Approximation of Translation Matrices
Z-Axis Translation Matrices
Interpreting Translations
Slices
Distance Measures for Spherical Harmonics
Integral Squared Error
Amplitude Spectrum
Bispectrum
Distance Measures for Real Spherical Harmonics
Implementation Details
Sampling Points
Fast Fourier Transform
Non-Spherical Input
Further Improvements
Noise
Image Preprocessing
Tangent Distance
Localization
Introduction
Experimental Setups
Method
Skyline Extraction
Scene Descriptors
Sequence SLAM
FABMAP
Datasets
Results
Precision versus Recall Plots
City Dataset with Tilt Variation
BMX Track Dataset
Disposal Site Dataset
Tilt-Invariance versus Sequence Length
Comparison of the Amplitude Spectrum and Bispectrum
Discussion
Conclusion
Holistic Visual 3D Compass
Introduction
Visual 3D Compass
Exhaustive Search
Rotation Parameterization
Fast Z/Y-Axis Rotations
Coarse-to-Fine Search
Global Illumination Invariance
Linearization of the Compass Search
Search Spaces
Parameter Sets
Experiments
Vanishing Points, Optical Flow, and Feature-Based Methods
Results
Single-Database Tests
Cross-Database Tests
Influence of Camera Translation
Feature-Based Methods on Raw Images
Discussion
Conclusion
3D-Warping
Introduction
Introduction to Warping
3D-Warping
Experiments
Parameter Sets
Results
Discussion
Conclusion
Overall Summary, Discussion, and Future Work
Summary
Skyline Segmentation
Spherical Harmonics
Localization
Holistic Visual 3D-Compass
3D-Warping
Discussion
Alternative Approaches for Skyline Extraction
Biological Plausibility
How Low Can You Go?
A Special Case: Movement in the Plane
Future Work
Multi-Snapshot Model
Robot Experiments: Proof of Concept
Robot Experiments: Lawn-Mowing
Conclusion
Proofs
Calculation Rules for Direct Sums and Kronecker Products
Clebsch-Gordan Matrix Ordering
Real Point-Wise Product
Symmetry Theorem
Rotation Theorems
Z-Axis Rotations
Y-Axis Rotations
X-Axis Rotations
Rotations of 90deg
Translations
Bispectrum for Real Spherical Harmonics
UVG: Details
HDR Algorithm Modifications
Efficiency on Generalized Data Sets
Numeric Stability
Spherical Harmonics & Applications
Code: Computation of Clebsch-Gordan Matrices
Detailed Results: Visual 3D Compass
Code: 3D-Warping
Full-Spherical Panoramic Image Databases
Experimental Setup
Database Descriptions