I have some ideas for algorithms, which would enable a more efficient search.
Use of pattern recognition could help with widespread areas, colour-matching and filters could ease the comparison of the before-after maps, or e.g. subtracting before-after maps could help identify changes more easily.
Is it possible to contribute to the effort with computational methods?
llja is the kind of person I would want to sit next to in class!
We recently published a couple blog posts about our use of algorithms; you might find them interesting
- Labeling at scale with smart feature prioritization
- Helping AI Algorithms Build Object Aware Maps of The World
We often incorporate the use of image algorithms into Tomnod projects. Sometimes the Tomnod community even contributes to campaigns for the purpose of improving image mining algorithms.