## Proceedings 1994 IEEE Computer Society Conference on Computer Vision and Pattern Recognition: June 21-23, 1994, Seattle, Washington |

### From inside the book

Results 1-3 of 79

Page 336

5 Incremental Surface

intersect the visible rim at the initial viewpoint. To achieve this, we incrementally "

grow" the ...

5 Incremental Surface

**Reconstruction**The goal of the global surface**reconstruction**task is to reconstruct the reconstructible surface regions thatintersect the visible rim at the initial viewpoint. To achieve this, we incrementally "

grow" the ...

Page 337

The following theorem gives the three rules the observer must obey for the semi-

global curve

rules) Let a be a finite-length curve drawn on the surface. If a intersects the visual

...

The following theorem gives the three rules the observer must obey for the semi-

global curve

**reconstruction**task. Theorem 2 (Semi-global curve**reconstruction**rules) Let a be a finite-length curve drawn on the surface. If a intersects the visual

...

Page 338

Incremental

points uniformly on the boundary of the reconstructed region Rule 4: Always

execute the Degenerate Patch

type of ...

Incremental

**Reconstruction**Behavior Rules that must be obeyed Rule 1: Selectpoints uniformly on the boundary of the reconstructed region Rule 4: Always

execute the Degenerate Patch

**Reconstruction**Behavior independently of thetype of ...

### What people are saying - Write a review

We haven't found any reviews in the usual places.

### Contents

Unifying Line Processes and Robust Statistics | 15 |

Object Recognition 2 | 23 |

Object Recognition 1 | 30 |

Copyright | |

86 other sections not shown

### Common terms and phrases

affine affine transform algorithm analysis angle approach approximation boundary camera completion components Computer Vision constraints coordinates corners corresponding curvature curve defined deformation derivatives described detection diffuse reflection direction dynamic edge eigenface eigenspace eigenvectors equation error extracted face filter frame function Gaussian Gaussian curvature geometric global hypotheses IEEE illumination inliers input invariant iteration kernel Lambertian reflectance light source linear matching matrix measure mesh method minimize motion estimation nodes noise normal object recognition obtained optical flow optimal outliers parallelogram parameters performance pixel plane pose primitives problem projection pyramid reconstruction recovered reflectance map regions representation robust rotation RPSDs scale scale-space scene segments sequence shape shown in Figure shows smooth solution space spatial specular specular reflection spine stereo structure superquadrics surface normal surface orientation surface reconstruction technique tion transform values vector velocity vertical viewpoint visible rim visual zero