| Introduction | p. 1 |
| The Problem | p. 2 |
| Applications | p. 2 |
| The Computer Graphics Approach | p. 5 |
| Avoiding the Model | p. 6 |
| A Review of Stereo Vision | p. 8 |
| Camera Model and Image Formation | p. 8 |
| Stereo Geometry | p. 10 |
| The Correspondence Problem | p. 13 |
| The Epipolar Constraint | p. 14 |
| A Simple Stereo Geometry | p. 16 |
| Rectification | p. 17 |
| Example: SSD | p. 19 |
| Contributions and Outline | p. 22 |
| A Survey of Image-Based Rendering and Stereo | p. 23 |
| Image-Based Rendering | p. 23 |
| View Synthesis Based on Stereo | p. 24 |
| View Interpolation | p. 27 |
| Mosaics and Layered Representations | p. 29 |
| Stereo | p. 32 |
| A Framework for Stereo | p. 32 |
| Preprocessing | p. 33 |
| Matching Cost | p. 33 |
| Evidence Aggregation | p. 34 |
| Disparity Selection | p. 35 |
| Sub-Pixel Disparity Computation | p. 35 |
| Diffusion-Based Techniques | p. 36 |
| Other Techniques | p. 37 |
| Promising Recent Approaches | p. 37 |
| Computer Vision Books | p. 38 |
| View Synthesis | p. 41 |
| Geometry | p. 42 |
| Three-View Rectification | p. 42 |
| The Linear Warping Equation | p. 44 |
| Computing the Rectifying Homographies | p. 45 |
| Synthesizing a New View | p. 47 |
| Resolving Visibility | p. 47 |
| Holes and Sampling Gaps | p. 47 |
| Combining Information from Both Images | p. 48 |
| Adjusting Intensities | p. 49 |
| Filling Holes | p. 50 |
| The View Synthesis Algorithm | p. 51 |
| Limitations of the Approach | p. 52 |
| Experiments | p. 53 |
| Image-Based Scene Representations | p. 60 |
| Summary | p. 61 |
| Re-evaluating Stereo | p. 63 |
| Traditional Applications of Stereo | p. 63 |
| Automated Cartography | p. 64 |
| Robot Navigation | p. 64 |
| 3D Reconstruction | p. 65 |
| 3D Recognition | p. 66 |
| Visual Servoing | p. 66 |
| Full vs. Weak Calibration | p. 67 |
| Comparison of Requirements | p. 68 |
| Stereo for View Synthesis | p. 68 |
| Accuracy | p. 69 |
| Correct vs. Realistic Views | p. 71 |
| Areas of Uniform Intensities | p. 72 |
| Geometric Constraints | p. 73 |
| Interpolated Views | p. 75 |
| Extrapolated Views | p. 77 |
| General Views and the Aperture Problem | p. 79 |
| Assigning Canonical Depth Interpretations | p. 80 |
| Does Adding More Cameras Help? | p. 80 |
| Partial Occlusion | p. 81 |
| Summary | p. 85 |
| Gradient-Based Stereo | p. 87 |
| Similarity and Confidence | p. 88 |
| Displacement-Oriented Stereo | p. 89 |
| The Evidence Measure | p. 90 |
| Comparing Two Gradient Vectors | p. 91 |
| Comparing Gradient Fields | p. 93 |
| Computing Gradients of Discrete Images | p. 94 |
| Accumulating the Measure | p. 96 |
| Experiments | p. 97 |
| Observing E¿ for Interesting Displacements | p. 98 |
| Stereo: 1D Search Range | p. 98 |
| General Motion: 2D Search Range | p. 105 |
| Computing Disparity Maps for View Synthesis | p. 105 |
| Occlusion Boundaries | p. 107 |
| Detecting Partially Occluded Points and Uniform Regions | p. 109 |
| Extrapolating the Disparities | p. 109 |
| Efficiency | p. 109 |
| Discussion and Possible Extensions | p. 110 |
| Summary | p. 111 |
| Stereo Using Diffusion | p. 113 |
| Disparity Space | p. 114 |
| The SSD Algorithm and Boundary Blurring | p. 116 |
| Aggregating Support by Diffusion | p. 121 |
| The Membrane Model | p. 122 |
| Support Function for the Membrane Model | p. 123 |
| Diffusion with Local Stopping | p. 125 |
| A Bayesian Model of Stereo Matching | p. 127 |
| The Prior Model | p. 127 |
| The Measurement Model | p. 129 |
| Explicit Local Distribution Model | p. 131 |
| Experiments | p. 134 |
| Discussion and Possible Extensions | p. 141 |
| Summary | p. 144 |
| Conclusion | p. 145 |
| Contributions in View Synthesis | p. 145 |
| Contributions in Stereo | p. 146 |
| Extensions and Future Work | p. 146 |
| Bibliography | p. 149 |
| Table of Contents provided by Publisher. All Rights Reserved. |