Getting started – Processing Bing Aerial Images for OpenStreetMap

So it’s time to get going with my project to build an automated process of identifying roads from Bing Aerial Images, I’ve written about it once before. I’ve now started to look into it a bit more. The OpenCV project seems to be a great starting point to work from, hopefully most of the code that I need to process images is already there so I can work on my application specifics. Before starting on your own it’s wise to search the web for similiar projects and I’ve found a few related sources.

I’ve two paths I need to go, one is to create a prototype to see if the Efficient Graph-Based Image segmentation works for my application and the other is that I need to write an report about the math behind it in three weeks. At the beginning of the week my goal was to get going with the report but the more I think about I feel that I will get a better understanding of it if I start with the prototype. I will write another blog post before the end of the week with more information.