Path finding is based on graph theory in which every viable/usable/possible/allowable position to be is represented by a node and possible routes between the nodes are called edges. In your question you state that white pixels are possible nodes while black pixels are no-go areas.

Lets assume that you have a small image 200x200 pixels and that 50% of the area is white and 50% black. That means there are 200*200/2 white pixels or 20,000 nodes and each node can connect to 8 others (thats with diagonal moves but ignores borders with black pixesl) thats 160,000 edges, lets reduce that to compenstae for borders with black pixels to 100,000 edges That is a MASSIVE graph to navigate.

Compare it with the Dijkstra example from the library (right). It has just 100 nodes and therfore <800 edges.

What I am saying is that you might want to consider what you are going to use as nodes.

I have never used this library because I have my own path finding library.

You might have a look at that, not because I claim it is better, rather to see whether it is easier for you to use.

Thanks for the link to your library. Even though I had trouble following the documentation (this is more of a reflection of my limited knowledge of object-oriented programming than a fault of your documentation), the example you provide is extremely helpful. I'm still not completely there with respect to integrating your library into my code, but I'm getting close. Again, your example code has been great for this.

Quark - thanks so much for your truly excellent library! The examples were incredibly helpful, and I was able to move forward on my project. (described in more detail here)