Taking photographs for raster to vector conversion

Unfortunately, using raster to vector software to vectorize the outline of an object on a photo is not as simple as it sounds. One reason for this is that before you can vectorize the object outline you need to divide the photo into two discrete color areas: The object and the background. This is usually hard to do automatically because the object and the background tend to share colors that make it difficult or impossible to separate them.

If you just take any old snap of an object, like the one of the baby below, you won't be able to separate the object from the background. In this instance it is quicker to simply trace around the outline you want manually.
 

Object and background share colors
There is too much commonality between the colors in the object and the background. It is impossible to separate them.

 
The trick is to photograph the object against a background which is a different color to the color of the object.

I got someone to hold a sheet of black card behind the baby while I took a photo. Now, the background and the baby are largely different colors and it is possible to separate them easily.
 

Object and background different colors   Object and background separated
The object and the background are different colors.
 
  They can be separated into two discrete color areas …
Vectorization of object outline    
… and vectorized.