demo https://omancoding.github.io/rihal_ocr/ (works in Chrome and Safari)
The tester can upload an image by drag and drop, the javascript will upload it to an endpoint which will return a text. The text is then fed to a function which will mainly use regex to create the rules that satisfies the specifications required. It will consult other services for certain information. For example, how does a word that start with capital letter will be identified as a city or a name or a landmark? Thus the usage of endpoints that query databases, which can be built by time.
Test images
