Is anyone aware of a site that offers bulk download of public domain images? I'm working on an image processing machine learning project, and I need a ton of photos to train and test my algorithm with. I've found several sites offering free public domain images, but they're all simple webpages, requiring you to navigate to each image in order to download it. I could potential write a script to crawl and download all images, but I'm hoping someone knows of a similar resource that simply has images available via an FTP share, ZIP archive, etc.
closed as off topic by mattdm, John Cavan, ahockley, Itai, chills42 Mar 29 '11 at 14:44
Questions on Photography Stack Exchange are expected to relate to photography within the scope defined by the community. Consider editing the question or leaving comments for improvement if you believe the question can be reworded to fit within the scope. Read more about reopening questions here. If this question can be reworded to fit the rules in the help center, please edit the question.
You can use wget to easily retrieve images from any web page or recursively from a hierarchy of web pages. So you need not restrict yourself to FTP sites, although wget works fine with FTP sites as well.
See the wget manual.
This page describes how you can filter by any type of file you wish to download.
You will in all probability get a few extra graphics files used for page decoration that you will have to manually screen out.
On a more general note, if you only use the images for training your system and don't publish, re-distribute or create derivative works then you can use pretty well any image on the Internet under the fair use interpretation of copyright law. Merely viewing the image already creates a copy on your hard disk. Now this note is sure to incite a few dissenting voices and we will have an interesting debate. Use your own judgment.
Even if you found a source, how'd you ascertain that the images are indeed free to download? I've found a lot of sites offering icon collections for download for software development for example that claim to be public domain but I had to discard them all because the actual icon collections contained a lot of icons stolen from existing applications (or containing trademarked images). I'd expect any collection of "free" images to be no different.
At best you can try to contact some photographers through their websites and ask for permission to use some of their imagery for your test.