Background: |
Representing real life provides an interesting challenge for us. Computers are very good at working
within finite spaces, bounded by digital data representations, computational precision, storage space, and network
transfer times. Life in general, and many structures within it, are continuous in nature, rather than discrete.
Finding the proper balance between realistic representation (modeling, rendering, etc.) and computational
tractability is one of the major threads flowing through all work in computer graphics. We often say: "If it
looks good enough, then it is good enough."
So, how does one go about representing real scenes in the computer? What options do we have? What are the
trade-offs? This project tries to show you how to address these questions, but does not ask you to answer
them.
|
Deliverables: |
Due Date: |
Activity: |
Results: |
Sep. 09 |
Field Trip: Fuller Labs Atrium, Salisbury Facade, Library Grove, Salisbury Pond |
|
Sep. 16 |
PART 1 Submission: Take three digital photos of scenes you find interesting, and submit them, along with a brief
explanation of why they are interesting. At least one of the images must be of an outdoor scene, and at least one
must be of an indoor scene. Please use a "decent" digital camera (not a cell-phone camera). If
you don't have access to one, you may check on out from the ATC on the first floor of Fuller Labs for free.
To submit your work, name the files something meaningful (e.g., "Lindeman_Outdoor1.jpg"), ZIP up all the files into an archive
called "LastName_FirstName_proj1-part1.zip", and email the ZIP file to me.
|
Original Pictures |
Sep. 23 |
PART 2 Submission: Choose one of your digital images, and deconstruct the scene. Identify:
- what you think is in the foreground, the middleground, and the background,
- what objects are the focus of the scene,
- what environmental properties are contained in the scene,
- what would be easy to model (not render) in the scene,
- what would be difficult to model (not render) in the scene.
Please limit yourself to no more than two pages of deconstruction. Include the image you have chosen in your
submission.
To submit your work, name the files something meaningful (e.g., "Lindeman_Deconstruction.doc"), ZIP up all the files into an archive
called "LastName_FirstName_proj1-part2.zip", and email the ZIP file to me.
|
Selected Pictures |
Oct. 07 |
PART 3 Submission: Create a simple mock-up of your chosen scene using only OpenGL. You can do this by:
- placing your virtual camera in the scene,
- placing translated, rotated, and scaled 3D primitives in the approximate locations of major scene objects
(you will be amazed by the expressive power of the mighty cube!)
- choose approximate colors for the objects.
Please submit a ZIP file with:
- The picture you are reproducing,
- A screen shot of your mock-up scene,
- The source file(s) and Makefile that produced your scene for MinGW or CCC, and
- A brief description of your work, including how to compile and run.
To submit your work, ZIP up all the files into an archive called "LastName_FirstName_proj1-part3.zip", and email the ZIP file to me.
|
First Mockups |
Oct. 28 |
PART 4 Submission: Create a more-realistic representation of your chosen scene using only OpenGL. You can do this by:
- replacing the simple 3D primitives with more-sophisticated models, consisting of multiple 3D primitives combined
together, and/or triangle meshes,
- adding more objects to the scene,
- refining your colors for the objects,
- getting more of the background environment modeled.
Please submit a ZIP file with:
- The picture you are reproducing,
- A screen shot of your final scene,
- The source file(s) and Makefile that produced your scene for MinGW or CCC, and
- A brief description of your work, including how to compile and run.
To submit your work, ZIP up all the files into an archive called "LastName_FirstName_proj1-part4.zip", and email the ZIP file to me.
|
Second Mockups |
|