<< 1 >>
Rating: 
Summary: Very hepful
Review: Anyone who is involved in areas such as computer graphics, computational radiology, robot vision, or visualization software should have a copy of this book. The author has done a fine job of introducing the most important algorithms in computational geometry, choosing the C language for their implementation. The choice of C might be somewhat dated now, since C++ is now beginning to dominate computational geometry, but readers who are actually programming these algorithms using C++ can easily extend the ones in the book to C++. Not all of the algorithms in the book are implemented into C, unfortunately, but the clarity of presentation is done well enough to make this implementation a fairly straightforward task. My interest in the book came from a need to design and implement algorithms for polyhedra in VRML and toric varieties in algebraic geometry. This book, along with others, was a great help in that regard. The running time of these algorithms was not really an issue with me, so the detail the author spends on discussing the complexity of the algorithms was not a concern. Readers who need to pay attention to running-time issues will appreciate his discussion of them for the algorithms that are presented. The ability to visualize objects in an abstract subject like algebraic geometry boils down to, in the case of toric varieties, to a consideration of how to manipulate polytopes geometrically. A major portion of the book, if not all of it, is devoted to the computational geometry of polyhedra. Because it is an introductory book, some more advanced topics, such as Bayesian methods to find similarities between polyhedra, and neural network approaches to classifying polyhedral objects are not treated. Readers who need to do such things will be well-prepared for them after a study of this book. In addition, there are good exercises assigned at the end of each chapter, so the book could be used in the classroom. Some readers will however choose to use it as a reference source, and it would be a good one, for the author gives references to topics that he only touched upon in the book. Some particular areas that were treated especially well were: 1. The discussion on data structures for surfaces of polyhedra. Although not very general, since he choose to deal with only triangulated polytopes, readers who need to be more general will have a good start in this discussion. 2. The discussion on volume overflow and how to deal with it using robust computation. 3. The discussion, albeit short, of the randomized incremental algorithm. 4. The treatment on the minimum spanning tree and Kruskal's algorithm. Communication network performance optimization is now a major application of this algorithm and others in graph theory, including the author's later discussion of Dijkstra's algorithm.
Rating: 
Summary: my rewiew
Review: i think that these website is very.it has everything that i need. all of my books are from amazan.
Rating: 
Summary: Response to Gonzalez review
Review: In his review Gonzalez says "the first exercise refers to 'every point in dP,' presumably meaning just the corner points (otherwise the problem would be unsolvable)." I did mean every point of the boundary of P, and the problem is solvable.
Rating:
Summary: A clear, concise text on fundamental Computational Geometry
Review: O'Rourke's approach reflects the essence of both "Computational Geometry" and the "C language" --- concise yet profound. The book covers the core subjects of Computational Geometry: polygon partitioning, convex hulls, Voronoi diagrams / Delaunay triangulation, "arrangements" of lines, geometric searching, and motion planning. The book assumes some familiarity with the C language, but is very readable even for non-C programmers. This is an excellent text for use as an introduction to Computational Geometry, a primer for Preparata & Shamos, while at the same time it's an excellent addendum to that more seminal text. By weaving working code into his presentation, O'Rourke gives traction to the powerful engine of Preparata & Shamos.
Rating:
Summary: A clear, concise text on fundamental Computational Geometry
Review: O'Rourke's approach reflects the essence of both "Computational Geometry" and the "C language" --- concise yet profound. The book covers the core subjects of Computational Geometry: polygon partitioning, convex hulls, Voronoi diagrams / Delaunay triangulation, "arrangements" of lines, geometric searching, and motion planning. The book assumes some familiarity with the C language, but is very readable even for non-C programmers. This is an excellent text for use as an introduction to Computational Geometry, a primer for Preparata & Shamos, while at the same time it's an excellent addendum to that more seminal text. By weaving working code into his presentation, O'Rourke gives traction to the powerful engine of Preparata & Shamos.
Rating: 
Summary: okay content, mediocre presentation
Review: This book provides a reasonable introduction to the field of computational geometry, although the notation is sometimes sloppy and the author frequently makes inconsistent assumptions about the reader. For example, on the first page he refers to a circle as a "one-dimensonial set of points," which although valid from a toplogical perspective is a little confusing in an introductory text. As another example, the first exercise refers to "every point in dP," presumably meaning just the corner points (otherwise the problem would be unsolvable). The book also sets up a lot of irrelevant mathematical definitions that generally obfuscate the presentation rather than clarifying it. Although not prohibitive for the ambitious reader, these needless hindrances are at best a little annoying. Secondly, I must criticize the text's scope, in light of the important role computational geometry has played in modern computer graphics. There is no discussion of clipping, culling, occlusion (e.g. BSP, octree, OBB), or even non-polygon primitives -- important topics arguably more useful to the target audience than e.g. convex hulls (to which over 1/4 of the book's pages are devoted). Regardless, this book (combined with a professor and a course) probably would serve quite well as an undergraduate text. Readers interested in a cookbook of applied graphics algorithms, however, should look elsewhere.
Rating:
Summary: Nice balance of theory with code
Review: This book was pleasantly surprising: I had expected to see code presented with minimal motivation or discussion of the underlying ideas -- something of a "Computational Geometry for Dummies" sort of book. That's not the case at all. This is a bona fide textbook on the subject, suitable for an undergraduate course.
It covers all of the the "classical" topics: convex hulls, line segment intersection, polygon triangulation, Voronoi diagrams, motion planning. The mode of presentation -- supporting a discussion of the theories with implementable code -- is actually a bit refreshing. For comparison: Other books, when discussing the line segment intersection problem (ie: Given a set of line segments, find all of their intersection points) simply assume that computing the intersection of a pair of segments can be done in constant time. This is not an especially difficult problem, but the discussion seems more complete with a brief description of how this might be done. The same can be said about other primitive tests and operations in other algorithms. Overall, this book can stand alone as an excellent introduction to computational geometry, but a serious student in the subject will want more: perhaps Preparata and Shamos or de Berg et. al.
Rating:
Summary: Nice balance of theory with code
Review: This book was pleasantly surprising: I had expected to see code presented with minimal motivation or discussion of the underlying ideas -- something of a "Computational Geometry for Dummies" sort of book. That's not the case at all. This is a bona fide textbook on the subject, suitable for an undergraduate course.
It covers all of the the "classical" topics: convex hulls, line segment intersection, polygon triangulation, Voronoi diagrams, motion planning. The mode of presentation -- supporting a discussion of the theories with implementable code -- is actually a bit refreshing. For comparison: Other books, when discussing the line segment intersection problem (ie: Given a set of line segments, find all of their intersection points) simply assume that computing the intersection of a pair of segments can be done in constant time. This is not an especially difficult problem, but the discussion seems more complete with a brief description of how this might be done. The same can be said about other primitive tests and operations in other algorithms. Overall, this book can stand alone as an excellent introduction to computational geometry, but a serious student in the subject will want more: perhaps Preparata and Shamos or de Berg et. al.
Rating:
Summary: Nice balance of theory with code
Review: This book was pleasantly surprising: I had expected to see code presented with minimal motivation or discussion of the underlying ideas -- something of a "Computational Geometry for Dummies" sort of book. That's not the case at all. This is a bona fide textbook on the subject, suitable for an undergraduate course.
It covers all of the the "classical" topics: convex hulls, line segment intersection, polygon triangulation, Voronoi diagrams, motion planning. The mode of presentation -- supporting a discussion of the theories with implementable code -- is actually a bit refreshing. For comparison: Other books, when discussing the line segment intersection problem (ie: Given a set of line segments, find all of their intersection points) simply assume that computing the intersection of a pair of segments can be done in constant time. This is not an especially difficult problem, but the discussion seems more complete with a brief description of how this might be done. The same can be said about other primitive tests and operations in other algorithms. Overall, this book can stand alone as an excellent introduction to computational geometry, but a serious student in the subject will want more: perhaps Preparata and Shamos or de Berg et. al.
<< 1 >>
|
