CS 350 - Spring 2009

CS 350 Spring 2009: Syllabus

The link to code/demos: here.
Sample grid data: here.
Visibility project material: here

Here is what happened in class.

Week 1 [Jan 20, 23]: Introduction to GIS. Intro to C. Intro to OpenGL.
Project 1
Materials:
- Intro.pdf | Pointers.pdf
- Emacs: Emacs reference card | Emacs quick reference
- The C programming language (wikipedia)
- The C Programming language
- The history of C
- C programming tutorial (by Steve Holmes)
- OpenGL: The Redbook : OpenGL Programming Guide (Chapters 1, 2, 3)
- Sample source code: graphics1.tar | graphics2.tar
Week 2 [Jan 27, 30]: Intro to databases.
Materials:
- IntroDB.pdf | Questions.pdf
- Readings:
  1. A relational model of data for large shared data banks (Codd, 1970)
  2. A good read for context on database models: What goes around comes around (Stonebraker, Hellerstein)
  3. General reading: The anatomy of a database system (Stonebraker, Hellerstein)
  4. Trivia: Edgar Codd | Codd's 12 rules
Week 3,4 [Feb 3, 5, 10, 12]: RDBMS and spatial data. Topological data representation.
Materials:
- spatialData.pdf
- Readings:
  1. DCEL (Half-edge) data structure:
    - The half-edge data structure (from Computer Graphics Algorithms FAQ)
    - DCEL data structure for 3d graphics (Bryan Holmes)
    - The half-edge data structure (Max McGuire)
    - Designing and implementing a half-edge data structure (Bronniman, WAE 2001)
  2. Quad-edge: Quad-edge data structures and library (Heckbert)
Week 5 [Feb 17, 19]: Virtual memory system. I/O-model and I/O-analysis.
Topics: Scanning. I/O-analysis of searching with a BST; quicksort, heapsort, mergesort. I/O-efficient sorting. I/O-efficient flow.
Materials:
- IOalgorithms.pdf
- Readings:
  1. Wiki: virtual memory
  2. Good blog on basic systems isssue (Gustavo Duarte)
Week 6, 7 [Feb 24, 26, March 3, 5]: Intro to geometric algorithms and fundamental geometric techniques: divide-and-conquer, line-sweep and incremental construction.
Topics: Area of a triangle, area of a convex and non-convex polygon. Orientation test. Collinearity. Betweenness. Segment intersection test. Point-in-polygon test. Convex hulls. Closest pair of points. Visibility. Segment intersection.
Speakers:
1. Seth Glickman: A general point-in-polygon test.
2. Will Richard: Convex hull. Incremental construction in O(n²). Improved incremental construction in O(n lg n).
3. Laura: Visibility on terrains. A brute-force quadratic algorith. A line-sweep algorithm for computing visibility on grid terrains.
4. Jack Morrison: A divide-and-conquer algorithm for finding the closest pair of points.
5. George Slavov: A divide-and-conquer algorithm for convex hull.
6. Jeremy Fishman: Finding the intersections between a set of vertical and horizontal line segments.
7. Tucker Hermans: Using A* to speed up Dijkstra SP algorithm.
8. Laura: Convex-hull lower bound (reduction sorting to ordered CH).
9. Laura: A line-sweep algorithm for finding the pairwise intersections of a set of line-segments in the plane (Bentley-Ottmann sweep).
Materials:
- introGeom.pdf
- SP animation: here
- Bentley-Ottmann sweep: demo
- Readings:
  1. Bentley-Ottmann sweep: notes (M. Schmid)
Week 8 [March 24, 26]: Visibility on terrains.
Speakers:
1. Laura: A different sweep algorithm for computing visibility on terrains. An I/O-efficient algorithm for computing visibility.
2. Mark McGranaghan: An O(lg n) approximation algorithm for vertex cover.
3. Guarding a terrain.
Materials:
- Readings:
Week 9 [March 31, Apr 2]: Line simplification. Terrain simplification.
Speakers:
1. Kyle Hebert: Douglas-Peucker algorithm for line simplification.
2. Imai-Iri algorithm for line simplification.
3. Surface simplification. Decimation. Incremental refinement.
Materials:
- gridvis.pdf
- Readings:
  1. Line simplification: Speeding DP simplification algorithm (Hershberger, Snoeyink) [see sec 1,2,3 for DP in O(n lg n)]
  2. Line simplification: A new approach to subdivision and simplification (de Berg, van Kreveld, Schirra) [see sec. 3 for Imai-Iri]
  3. Terrain simplification: Digital elevation models and TIN algorithms (van Kreveld)
  4. Terrain simplification: Fast polygonal approximation of terrains and heigh fields (Garland & Heckbert)
Week 10, 11, 12 [Apr 7, 9, 14, 16, 21, 23]: Visibility.
Materials:
- Readings:
  1. Approximating the visible region of a point in a terrain (Boaz Ben-Moshe, Paz carmi, Matthew Katz, Alenex 2004).
  2. Visibility preserving terrain simplification (Ben-Moshe, Joe Mitchell, Matthew Katz, Yuval Nir, SoSG 2002)
  3. Algorithms for visibility compytations on digital terrain models (Leila de Floriani and Paola Magillo)
  4. Representing the visibility structure of apolyhedral terrain through a horizon map (de Folriani, Magillo)
  5. Fast horizon computation on all points of a terrain with visibility and shading application (J. Stewart, IEEE Trans. Vis & CG, 1997)
  6. Computing teh approximate visibility map, with applications to form factors and discontinuity meshing (J. Stewart and Tasso Karkanis, 1998)
  7. Hierarchical visibility in terrains (J. Stewart, Eurographics Rendering Workshop, 1997)
  8. Smugglers and border guards - the geostar project at RPI A(Franklin, Inanc, Xie, Tracy, Cutler, Andrade, and Luk, ACMGIS 2007ACM GIS 2007)
  9. Multiple observer sitting and path planning on compressed terrain (Tracy, Franklin, Cutler, Andrade, Franklin Luk, Inanc, and Xie, SPIE 2007)
  10. Tradeoffs when multiple observer siting on large terrain cells ( Franklin and Vogt, SDH 2006)
  11. Efficient observer siting on large terrain cells (Franklin and Vogt, GIScience 2004)
  12. Approximating visibility (Franklin, GIScience 2000).
  13. Higher isn't necessarily better: Visibility algorithms and paper experiments (Franklin and Ray, SDH 1994).
Week 13 [Apr 28, 30]: Voronoi diagrams and Delaunay triangulations.
Materials:
- Demo: VoroGlide | Fortune's sweep
- Readings:
Week 14 [May 5]: Class summary. Invited talk.

Topics we did not cover:

TINs in GIS and TIN algorithms. Isoline extraction and the contour tree.
An I/O-efficient priority queue.
Visibility of arbitrary objects. The horizon-tree.
Edge-quadtrees and segment intersection (map overlay).
The art-galery problem (polygon guarding). Heuristics for polygon guarding.
Guarding 1D-polygonal chains. Heuristics.
A 2-approxomation for TSP using Delaunay triangulations.