I’ve put up a page to help get hold of pretty pictures of living organisms. I often need to do this when giving talks and trying to avoid text-heavy slides. One time I might need a rather general picture, such as a bat to discuss animal sonar. Other times I might need much a more specific image, such as a vampire bat nose (when talking about the nasal heat receptors of bats, and their evolutionary links to spicy heat).
The problem is that my talks might be filmed, stuck on a website, or whatever. Unless this can be argued as “fair use“, that technically rules out images that require attribution, such as CC-BY. That’s because even if I put a small attribution text on each slide, or collect the photographers’ names together at the end, this is unlikely to be preserved in a short, relatively low-resolution film of a talk. Pedantic, I know. But I generally try to err on the side of caution.
So the ideal images are ones in the public domain. Is it possible to search for public domain pictures of particular animal or plant groups, and (ideally) rank them by quality of the image? Google’s advanced image search can be restricted to images which are “free to use, share or modify, even commercially” – but that includes CC-BY images. The best option at the moment seems to be to use the Encyclopedia of Life. EoL not only provides a way to get at images classified by whatever level of taxonomy you care to choose (e.g. the 52 thousand images of various beetle specimens in the database), but also classifies images into cc-by, cc-by-nc, cc-by-sa, cc-by-nc-sa, pd [public domain], and na [licence not available]. Unfortunately, the EoL website doesn’t yet allow you to filter images based on licence type.
Currently, the only way I can see of doing this is to use the EoL API. I’ve just put up a page to do this at http://yanwong.me/EOLimagesearch.cgi. Alternatively, you can do a standard EoL search to find the numeric ID of the correct taxonomic group (e.g. searching for “beetle” leads immediately to a page with the ID 345), then feed that numeric ID into the EoL Page API test page, and root through the results to find the EoL data object IDs. Or for Mac users, my perl script at http://yanwong.me/?page_id=861 will bring up a web page with the first 75 images.
I’m not sure there’s a way to get any beyond that at the moment, but it’s a start. When I get the time, I’ll try to put up a web page that does this automatically.
Some caution is needed with the results, as not all images marked on EoL as public domain really are. But a test run suggests that about 10% of mammal species seem to have a public domain image available on EoL (many of them are old book illustrations).