Video Review :)
Monday 2 February:
- Interactive Boundary Computetion of Boolean Combinations of Sculptured Solids
Studies the model of a Bradley vehicle on an SGI, the problem of avoiding colisions.
Interactive design, real time very important ( used in CAD applications).
3D problem.
Key words: boolean operations, boundary representations.
- OBB-Tree
Deals with the possibility of obtaining smooth models. There were different
models shown, for example the engine-piston one. All are on SGI.
Key words: rejection set, colision query ( lots of rectangles there!:)
3D problem, performance and complexity very important as the alghorithm was
supposed to run in real time.
- walkthrough 96, Real Time Rendering of Massive Models
Definition of massive models (> 500 pieces), the problem is the time consumed
to render them. Solution: rendering only the noticeable parts ( in view) and
constructing smaller models.
3D problem.
Key words: the problem of Portals and Mirors, Dynamic Texture-Based Simplification,
Radiosity, Envelope Surfaces (simplifies the algorithm about 2-3 times, preserving
the quality), triangles.
- The 5th Annual Video Review of Computational Geometry
The (approximative) titles would be:
- Radiosity in Flat Land Made Visibly Simple
- Enumerating: Delaunay Flip
- Reverse Search
- Enumeration of Regular Triangles
- Four Polytops and a Funeral (for me)
- A Package: Triangulations
- Impulse-Based Simulation
...and what they were about:
- Radiosity, reflection, used for constant illumination, rendering immages.
Key-words: duality transform, visibility complex, the form factor, meshing, umbra,
penumbra.
3D problem.
- Testing central systems through direct simulations of impulses.
Colision symulation, 3D problem.
Key -words: Culling checks, spatial partition, multibodyes, convex hulls, rotation
diagram, placement poligon, power of coherence, parametric surfaces.
View problem, colision detection.
Uses Delaunay triangulations.
- Polytops, a 3D versus 2D problem, define and deform a polytop.
Contains minimum weight edges problem -> minimum weight triangulations.
Problems with the complexity of the algorithm.
- Triangulations, min weight (total length of edges), Markov chain and random
generation. Uses the adjacency graph.
Wednesday 4 February:
Two tapes:
- The 4th Annual Video Review of Computational Geometry
- Hip Air
3D problem, studies the motion in mechanism.
Key -words: contact ( constrains moving), configuration space, toleracing.
Problems: jamming, multiple mechanisms, Fuji camera.
- 3D Modelling-Delaunay Triangulations
Studies the reconstruction of a bone-shape from magnetic images of its sections.
Key-words: contour, Voronoy diagrams, triangulation.
2D versus 3D ( through projections).
- Incremental Collision Detection for Polygonal Models
Polytops, convex-non convex objects.
Colision problem, tori simulated motion.
Key-words: Depth First Search, Voronoy diagram.
- Convex Surface Decomposition
The problem of boundary selection and decomposing of boundaries of real-life
objects.
3D problem.
Key-words: convex hall, Breath First Search, minimal surface decomposition.
- Visibility Complex, not enough time:), lots of projection in 2D.
- Animation of Euclid Proposition
- The 6th Annual Video Review of Computational Geometry
- The Bisector Surface of Freeform
Bisector plans.
Key-words: rational space curves, planar versus 3D curves, B&ecute;zier curves.
Nice images.
3D Visibility Made Visibly Simple ( again:)
An introduction to the Visibility Skeleton, umbra, penumbra, graphs associated
the visibility problem.
The room model.
Key-words: light source, visibility, umbra, penumbra.
Application of an Effective Geometric Clustering Method to Color Quantization Problem
Key-words:Clustering problem, minimazing the cost, K-clustering.
Aplication in photo-modifications, quality.
Animating the Offset Poligon.
Distance function, robot localization, local offset objects.
Key-words: Voronoy diagram, scaled polygons versus offset polygons ( they
shrink till they become a point).
Useful for finding the Voronoy diagram.
Triangulation-Based Object Reconstruction Methods.
Approx 3 steps:
- Point sampling, 3D Delaunay, A-alpha-shapes->
- dense triangle mesh, mesh simplification, base mesh->
- A-patch fitting, C one smooth mashed.
It has several advantages, but disadvantages too.
- Approximating Weighted Shortest Paths on Polyhedral Surfaces
Key-words: Polyhedral terrain, weight-cost, shortest cross path.
Problem: finding paths on weighted terrains, algorithmic complexity.
Several approaches, first associate a graph to the terrain ( bad approx).
Improvement -> add steiner points to the graph, fine tuning ( unweighted),
with linear time, fine tuning ( weighted), more precise, but more complex too.
Class Home Page
My Home Page
© 1998 Octavia Petrovici