Text extraction results on a set of photographs of business opening hours generously shared by Lukas Toggenburger on the OpenStreetMap Talk mailing list. The photos were run through Amazon Rekognition's text extraction feature and the results saved as JSON files.
The source photographs are much too large to be hosted on GitHub (12.5GB), but at the time of writing they are available at this link, provided generously by Bryce Cogswell.
This repo contains text recognition results from the "deduplicated" folder in this photo set. The name of each JSON file corresponds to the name in the photo set.