All test resources must come from free public domain sources for copyright reasons. Test files do not necessarily produce perfect (or even good) OCR results. +-------------------+--------------------------------------------------------------------------------+ | File | Source | +===================+================================================================================+ | graph.pdf | Wikimedia | +-------------------+--------------------------------------------------------------------------------+ | c02-22.pdf | Project Gutenberg: https://www.gutenberg.org/files/76/76-h/images/c02-22.jpg | +-------------------+--------------------------------------------------------------------------------+ | LinnSequencer.jpg | Wikimedia: https://upload.wikimedia.org/wikipedia/en/b/b7/LinnSequencer_hardware_MIDI_sequencer_brochure_page_2_300dpi.jpg | +-------------------+--------------------------------------------------------------------------------+ | congress.jpg | http://www.baxleystamps.com/litho/meiji/courts_1871.jpg | +-------------------+--------------------------------------------------------------------------------+