• View
• History
• Print

Alex / Project1Proposal

Project 1 Proposal:

1. A description of the project

Implementing a mesh data structure in object-oriented C++ style from scratch.

2. A discussion of how the project scales

Once a mesh is in place there are several operations that I'd like to develop on it, including cutting, smoothing, and hopefully subdivision.

3. A plan of execution

• Step 1: Load, View, Save .PLY files
• Step 2: Robustness testing. Loading several more complicated models and making sure its integrity is verified. For example, testing certain invariants like 'The twin of every edge's twin is the original edge'.
• Step 3: Implementing operations on the mesh. Probably smoothing first, then something like cutting, and then subdivision.

4. A description of where you need to be at the milestones

• At Milestone 3 I should have the load/view/save functionality implemented and stable. I'm actually almost there.
• At Milestone 4 I should at least have smoothing implemented, and I should at least have some work done towards implementing cutting.
• After that it's a matter of making sure the mesh retains integrity after smoothing and cutting, then implementing a subdivision scheme.

5. An acknowledgment of the risks, preferably with some fallback

The biggest risk when implementing any of this is that the mesh loses its structural integrity (eg loses connectivity) at some point.

6. A description of what you expect to have in the end, and a description of how you will evaluate your results.

At the end I will have an application that will load and display a PLY model, as well as manipulate it in various ways, and save it to a file.

7. A description of what tools you will use, what resources you will need, etc.

I'm basically hijacking your ShaderAssignment project from 559 last semester as an application framework. It's very simple, integrates an arcball cam for viewing, and is easily extendable to handle other events.

I'll be coding in Visual Studio 2005 in C++.

8. A description of what you expect to learn from doing the project.

Practice developing new data structures is always helpful as a programmer, I think. I'll also get a really good foundation is mesh processing and manipulation. The faster I can get the general data structure functioning well, the more I'll be able to learn about different manipulation techniques.