Here is a great beginner's tutorial for it assuming you have some knowledge of browser languages:
There are two resources by Publishing Houses, that provide image search in articles.
One from Springer - SpringerImages.
Second is from Cell - CellImageLibrary.
Of course they do have bioinformatics images.
If using python, consider the seaborn package (gallery here) which is built on top of matplotlib. It includes heatmaps, many statistical plots including linear models, mult-faceted/grid plotting, and color palettes (including colorbrewer). It leverages the pandas DataFrame object to make it easier for plotting then standard matplotlib. Personally, I like the figure aesthetics more than ggplot2.
For color palette selection of charts, I use: IWantHue "Colors for data scientists"
They do k-means clustering of colors in a selected range to achieve maximum divergence between the individual colors. However, you can keep colors and re-cluster and similar thing to make it suit your needs.
If you want to create animations or explore high-dimensional data, GGobi is the tool to use. You can use it in R (package "rggobi") which is a great convenience. I would highly recommend a visit to r-forge, where you can search through lots of interesting projects with a wide range of bioinformatics applications.
For general plotting, you should consider GD, though you probably already have it if you generate graphics in BioPerl. Since you mentioned Perl, there is a great list of relevant modules on CPAN.