Vectorization: Hands-on experience!

Hands-down, the first step to learning vectorization should be hands-on.

Unfortunately, I lack the following (which are often used for vectorization):

• Fancy tablet pen
• Vectorization software

Of course, the second thing is easy to get online for free; there’s resources like InkScape, for instance. But I felt lazy, so here’s what I restricted myself to:

• Mouse
• Microsoft Powerpoint

As it turns out, Powerpoint 2007 is a pretty fancy piece of software. You know that “curve” tool? Yeah, that thing is pretty useful. I used it to trace over the following picture, because ‘Murica:

At first, the picture looked kind of boring, since I just captured the most prominent features. This could be considered the “first iteration”:

But hey, it seems like it’s getting somewhere. So I added a few more “Smart Shapes” and colors, and glasses:

Eureka! Fabulous! So next time you want to draw something cool, look no further than PowerPoint. If you’re interested, the original file is here. (Use group -> ungroup as necessary to get the individual shapes.)

A New Project: Vectorization

Most of our images on the internet are raster, or “pixel,” images. That means they can’t be scaled – when you zoom in, they’ll become fuzzy. So how can we avoid this? Vectorization.

Vectorization is the process of defining images by their geometric relationships with each other. The text on your computer is an example – you can zoom in forever and the edges will still be smooth. Humans also perceive in vectors – we distinguish objects by their edges, not by their appearances on a fixed array. SVG’s, or Scalable Vector Graphics, are useful tools for creating great-looking scalable images; however, they always end up looking “cartoony” because of the limitations of computer shading. To see what I’m talking about, check out this world map – and try zooming in as much as possible: http://upload.wikimedia.org/wikipedia/commons/6/63/A_large_blank_world_map_with_oceans_marked_in_blue.svg.

So, for the next few weeks I’m going to find out how vectorization works – the algorithms, the philosophy behind it, etc. Hopefully, if this turns out good, I’ll be able to make my own vector images from raster images!