CSE681 Lab 2

Winter 2011
due Jan. 31


input file	input file

input file	input file

FINALIZED

Objective

To add

the use of a scene description file
the use of transformations (NOT including non-uniform scale)
handling multiple light sources.
handling multiple spheres.
modeling diffuse and specular illumination,

Assignment

Modify your ray tracer to incorporate new capatilities and effects. Call it lab2.
Lab2 should use the argument string as the input file (as was done in Homework #2). See below for the scene description file commands.
Lab2 should output a ppm file to a file named in one of the scene description commands.
You should accomodate mutliple spheres - up to 10.
You should accomodate multiple lights - up to 10.
Model diffuse and specular illumination.
Create and submit your own input files to demonstrate the features of your program.

Scene Description File

Each scene description file can contain any number of the following commands in any order.
Read and process the commands until an end-of-file is encountered.
- for this lab, ignore the aspect parameter - always assume it is '1' (square pixels)
- transformations are only applied to objects, not light sources (different from OpenGL)
skip over blank lines
allow one or more spaces between values
a command can span multiple lines
a command always starts at the beginning of a new line
colors are specified in the rangle [0.0:1.0]
When an end-of-file is encountered, compute the image and output it to the output file.
You should have default values, as specified in lab1, for the following commands: camera, ambient, background, and output.

command	action
#	Ignore all text up to the end-of-line
camera <x> <y> <z> <cx> <cy> <cz> <angle> <aspect> <hither> <yon> <tilt> <xres_i> <yres_i>	Record the camera parameters: position, center-of-interest, vertical half angle of view, aspect ratio, hither and yon clipping distances, tilt (rotation around viewing direction from 'up', x and y image resolution (integers).
ambient <a>	Record the ambient value.
background <r> <g> <b>	Record the background color.
output <filename_s>	Record the output filename.
identity	Set the current matrix to the identity.
rotate [x\|y\|z] <degrees>	Form the appropriate rotation matrix and multiply the current matrix on the left
translate <tx> <ty> <tz>	Form the appropriate translation matrix and multiply the current matrix on the left
scale <x> <y> <z>	Form the appropriate uniform scale matrix (for this lab, you can just use the 'x' parameter) and multiply the current matrix on the left
sphere <r> <g> <b> <texture-name_s> <kd> <ks> <n_i> <kr> <gl> <kt> <n1> <tr>	Create an additional sphere in the scene data structure, transforming it according the current transformation matrix and record its parameters: color, ignore the texture name text string, diffuse coefficient, specular coefficient, specular power and for this lab, ignoring the last five parameters.
light <x> <y> <z> <intensity> <size>	Create an additional light in the scene data structure and record its parameters: position, intensity, and, ignored in this lab, size.

Sample submission and execution

Submit the source code for your program. For example:

submit c681aa lab2 lab2.c makefile

Of course, you can break it up into multiple source files and submit those. For example:

submit c681aa lab2 main.cpp camera.cpp sphere.cpp light.cpp makefile

- just as long as the grader can make and execute your program using whatever input file the grader wants to as an argument to lab1:

make
lab2 <input filename>

and the ppm file should be found in whatever output filename is specified in the input file (or 'output.ppm' if none is give).

Under UNIX, the ppm file, output.ppm in this example, should be able to be viewed using:

convert output.ppm output.jpg
display output.jpg

Extra credit

Submit a README.txt along with your lab files to explain the extra credit you are applying for.

Additional object types (up to 10% depending on complexity)
- Display an ellipsoid, cone, cylinder or other type of object.
- Come up with an appropriate object command and ray intersection routine.
Background scene (5%)
- Process the scene description command:
```
backgroundFile  <filename>
```
- Using the filename, read in an image of the appropriate resolution and use as the background colors.

Notes

To parse the input file, your Homework #1 and #3 should provide the code to maintain the current matrix, populate the scene with lights and spheres, and set the camera parameters.
For the illumination model, use the following:
- Implement the basic illumination model using white lights with diffuse and specular illumination:
```
c = kd*(ambient + ∑(N⋅L_i)*I_i)*object_color + ks*∑I_i*(R_i⋅E)ⁿ*(1,1,1)
```
  where the summation is over the light sources that the surface of intersection is oriented towards (i.e., N⋅L_i > 0).
- If L_i⋅N < 0 then the point on the surface is facing away from this light source and should not contribute to the illumination at all.
- Also check if R_i⋅E < 0, then skip the specular component (thanks to Michael Schoenberg for pointing this out as a potential source of error)
- At a given pixel, if one of the R,G,B values is larger than 1.0, scale all the values so that the largest value is 1.0
- Don't clamp the low end; the ambient has already been added in.
For the sphere command, create a new sphere, store the sphere paramters, and record the current matrix. In later labs, you will also have to record the current inverse matrix.
suggested classes, attributes, primary methods:
- material: color, kd, ks, n, etc.
- intersection: point, normal, material
- sphere: material, transformation matrix, inverse transform matrix
  intersect(ray)
- light: position, intensity
- scene: spheres & lights
  getFirstObject()
  getNextObject()
- camera: all camera parameters
  takePicture(scene)

Last updated 1/26/2011

class home page