|
|
|
Creating 2D Images
May 2001
Learning something new is usually most difficult in the beginning. However,
once the initial fear is overcome and the basics are understood the learning
becomes easier. This is true for computer graphics. At first computer
graphics can seem a little overwhelming, but if you look at the basics it
isn't that bad.
The main topics to be discussed are...
This page is not devoted to revealing all the mysteries of programming computer
graphics, it is mearly an introduction to the basics. 3D computer graphics
will not be discussed, only the creation of 2D images. No special API like OpenGL is needed to run the code below, just
the standard C++ libraries. It is assumed that
the reader has an itermediate level of C++ programming experience. The code
provided below was written in C++ using classes and dynamic memory
allocation.
2D Images
It is important to understand how 2D images are generated before diving into
3D interactive scenes. Mainly because everything viewed with a computer
monitor is in 2D. Some images look 3D, but they are actually just 2D. Think of
the monitor as just one giant 2D image. The monitor is made up of an array
of pixels, and each is assigned a particular color. The colors are made up
of 3 different values, a Red component, Green component, and a
Blue component (RGB). Each component is simply an unsigned
char value in the range [0..255]. This is all the information needed to
create and write a color image in 2D!
The below code can be used to represent all the above using C++.
// used to contain color information about each pixel
class Pixel {
unsigned char red;
unsigned char green;
unsigned char blue;
Pixel ();
void setPixel (unsigned char r,
unsigned char g,
unsigned char b)
{
red = r;
green = g;
blue = b;
};
};
...
Pixel image[800][600];
The above is not the best way to represent an image, but it would work. Now
the image just needs to be saved in a format readable by an image viewer.
Writing PPM Files
One of the easiest formats to store images in is the Portable PixMap
(PPM) file format. It isn't readable by Netscape, unless the proper
plug-in is
installed. Kevin Kurtz wrote a
PPM viewer plug-in for Netscape on Windows and can be downloaded here. xv on the UNIX platform will read
PPM files.
A PPM file can be either
written in ASCII text or binary format. Better explainations for reading and
writing PPM files can be found by searching the internet, but below is a
quick example.
P3
# my PPM file
4 3
15
0 0 0 0 0 0 0 0 0
0 0 0 0 100 7 0 0 0
0 0 0 0 0 0 0 100 9
100 0 100 0 0 0 0 0 0
The header contains the keyword "P3" which signifies an ASCII PPM file, a
comment started with a '#' character, the image dimensions, and the total
number of colors the image contains.
Image Domains
Once individual pixels can be easily modified and an image can be written to
file, it is time to become slightly more advanced. Instead of referencing
each pixel by a pair of integer values it is better to create an image domain
using floating point values which are then mapped back to integer pairs. For
example, if the image
is 200x100 (pixels), then the image domain may be (0.0..2.0, 0.0..1.0).
Now an infinite number of points can be referenced from the image (of course
there are still a finite number of pixels). One of the
main goals behind creating these image domains is so create nicer images. This
might seem unnecessary now, but if this can be useful when creating
anti-aliased images.
Coordinate Systems
The coordinate system used the most in computer graphics is the
Rectangular system with X and Y axes. However, often in a 3D scene it is
necessary to apply a texture to a surface. If the surface is a sphere then a
Spherical system with with S (theta) and T (phi) axes can be used.
The Image Class
The Image class was written to create an image with a desired
domain and set points within the domain to a color.
The Image class contains public member functions to assign a color to
an individual point, or draw a dot centered at a point. When drawing a dot,
the user specifies the center point, color, radius and filter function to be applied to the dot.
Drawing dots instead of setting points often greatly reduces the number of
samples required to render a scene (provided the proper radius and filter function are chosen).
The drawdot Program
drawdot was written as a driver for the Image class. Its
standard usage with out any arguments will create an image with a single dot
at the origin and output a binary formatted image to
stdout. However, drawdot is also equipped with several
command line options.
Usage : drawdot [-file name] [-size x,y] [-color r,g,b]
[-point x,y,z] [-rad n] [-filter 1..5]
[-random num]
Options:
-file name : Name of file to save image to,
default=STDOUT
-size x,y : dimensions of image in pixels,
default=200x200
-color r,g,b : Color of dot to be drawn,
default=random color
-point x,y,z : Point in image domain to be drawn,
default=origin
-rad n : Radius of dot, same units as image
domain, default=0.5
-filter 1..5 : Filter to be applied to dot.
1 = BOX
2 = TENT
3 = QUADRATIC
4 = CUBIC
5 = QUARTIC
default=TENT
-random num : When this argument is set, num randomly
positioned, filtered, colored, and sized
dots are drawn.
Creating Animations
Creating an animation is really not complicated at all. An animation is just
a series of images displayed at a constant rate. With this basic
definition in mind the drawdot program can be used to create some
simple animations. drawdot was set up with command line arguments to
position a dot at any point within the image domain, along with any radius,
color, and filter. A script can very
easily be written to drive the drawdot program to generate an
animation.
Below are examples of gif animations with links to the Perl scripts used to create them. The first
is a single dot
bouncing within the image domain, the second is a single dot with its radius
expanding and shrinking, and the third is a single dot changing from red to
purple to blue and then in reverse. All of them were created by calling
drawdot a specified number of times, converting the images to gifs
using ppmtogif and then using gifmerge to generate the
animation. (ppmtogif and gifmerge are available on most UNIX
sytems and souce code can be found on Google relatively quickly.)
Code
All the code used to create the above images and animations is provided
below. Links to the individual files are available as well as a link to a
tar file containing all files and scripts.
|
