Nguyen, Xuan Vinh: Super-resolution compressed sensing for resolving time-of-flight multipath interferencesSuper-resolution Compressed Sensing für die Lösung von Time-of-Flight Mehrwegsinterferenz-Problem. 2017
Content
- Acknowledgments
- Abstract
- Zussamenfassung
- List of publications
- Contents
- List of figures
- List of tables
- List of abbreviations
- 1 Introduction
- 1.1 Time-of-flight imaging camera and multipath interference problem
- 1.2 Multi-frequency TOF acquisition using CS
- 1.3 Outline
- 2 Time-of-flight principle
- 2.1 Pulsed modulation
- 2.2 Continuous wave modulation
- 2.3 Main hardware components
- 2.4 State-of-art TOF problems and solutions
- 2.4.1 Measurement linearity error
- 2.4.2 Photon shot noise
- 2.4.3 Saturation
- 2.4.4 Ambiguity range
- 2.4.5 Multipath interferences
- 2.5 Summary
- 3 Compressed sensing and TOF multipath problem
- 3.1 Compressed sensing (CS) and sparse recovery
- 3.2 Sparse time profile of TOF multipath problem
- 3.3 Multiple frequency TOF (MFT) measurements
- 3.3.1 Mismatch model errors
- 3.3.2 Depth resolution of target discrimination
- 3.3.3 TOF modulation frequency limitation problem
- 3.4 Super-resolution MFT compressed sensing
- 3.4.1 Preliminary
- 3.4.2 High coherence of super-resolution sensing matrix
- 3.4.3 Minimum distance
- 3.4.4 Relaxed sparse support evaluation
- 3.4.5 Numerical analysis
- 3.5 Summary
- 4 Super-resolution compressed sensing methods
- 4.1 Band exclusion and local optimization
- 4.2 Modified cyclic orthogonal matching pursuit -OMP3
- 4.2.1 Global optimization
- 4.2.2 Modified cyclic OMP - OMP3
- 4.2.3 Combination between OMP3 and LO technique
- 4.3 Non-negative least squares optimization - POMP
- 4.3.1 Non-negative constraints
- 4.3.2 Negative atom removal module
- 4.3.3 Advantages and disadvantages
- 4.4 Non-negative magnitude adjustment orthogonal matching pursuit
- 4.4.1 Basic idea
- 4.4.2 Non-negative magnitude adjustment orthogonal matching pursuit - Ma-OMP
- 4.4.3 Modified non-negative magnitude adjustment orthogonal matching pursuit - Ma-OMP3
- 4.4.4 Optimized adjustment factor
- 4.5 Numerical results
- 4.5.1 Preliminaries
- 4.5.2 Optimized adjustment factor of Ma-OMP3
- 4.5.3 Comparison between different methods
- 4.5.4 Considered parameters
- 4.6 Combined OMP based on predicted minimum distance - CMD-OMP
- 4.6.1 Basic idea
- 4.6.2 Tuning parameter estimation
- 4.6.3 Minimum distance prediction
- 4.6.4 Numerical results
- 4.7 Summary
- 5 Multi-frequency selection optimization
- 5.1 Frequency selection optimization
- 5.2 Frequency and initial phase-offset optimization
- 5.3 Numerical results
- 5.4 Summary
- 6 Multiple measurement vector in super-resolution compressed sensing
- 6.1 Multipolarization TOF signal model
- 6.2 Multiple measurement vector (MMV) model
- 6.3 Conventional OMPMMV
- 6.4 Modified variants of greedy pursuits for MMV model
- 6.4.1 Modified OMPMMV
- 6.4.2 Modified global optimization - GO-MMV
- 6.4.3 Cyclic orthogonal matching pursuit for MMV model - OMP3-MMV
- 6.4.4 Cyclic magnitude adjustment orthogonal matching pursuit for MMV model - Ma-OMP3-MMV
- 6.4.5 POMP for MMV model
- 6.5 Numerical results
- 6.5.1 Comparison between OMPMMV variants
- 6.5.2 Comparison between OMP3-MMV variants
- 6.5.3 Comparison between Ma-OMP3-MMV variants
- 6.5.4 Comparison between MMVs and SMV
- 6.5.5 Number of MMVs
- 6.6 Summary
- 7 Simultaneous multiple frequency acquisition
- 7.1 Projection matrix model
- 7.2 Projection matrix optimization in SMF acquisition
- 7.2.1 Projection matrix optimization method
- 7.2.2 Modified projection matrix modification
- 7.2.3 SMF model parameters
- 7.3 Numerical results
- 7.3.1 Coherence histogram
- 7.3.2 Support recovery performance
- 7.3.3 Multiple-SMF acquisition using MMV techniques
- 7.3.4 Relaxed super-resolution problem
- 7.3.5 Summary
- 7.4 Joint multiple frequency calibration
- 8 Experimental results
- 8.1 PMD Multicam System and measurement matrix formulation
- 8.2 Transparent object imaging
- 8.3 Comparison of various reconstruction methods
- 8.3.1 POMP and OMP3
- 8.3.2 BPIC
- 8.3.3 Ma-OMP3
- 8.3.4 Supportive affects of LO techniques
- 8.3.5 Summary
- 8.4 CMD-OMP
- 8.5 Frequency and phase-offset selection optimization
- 8.6 SMF and joint multiple-frequency calibration
- 8.7 Summary
- 9 Conclusion
- 9.1 Super-resolution compressed sensing algorithms
- 9.2 Super-resolution MFT sensing matrix design
- 9.3 Relaxed super-resolution factor
- A Appendix
- Bibliography