[smc-discuss] ocr help ;)

[Mark]

hey saw on github you have a background on video analysis/ocr type stuff. and wanted to run my problem past you :D

i’m working with tesseract on some recognition on some video streams. i need assistance improving and also maybe looking at other image
recognition libraries (opencv). I have a number of streams that have a different elements in them, so each have to be designed for recognition differently.

So the current streams are over twitch. 1 of the issues is that sometimes twitch streams in different quality - I have 720p, 480p, 360p but there are two 480p streams with different UI position. and its how to handle that along with different game endings so what happens if goes to extra time/ penalties etc. what i need to know if
the winning team and the score, and also the gamertag of the person that won oner the top of the player names here are some of the
recorded game streams with descriptions of them…

https://www.youtube.com/watch?v=w7VQrMDw5-0

https://www.youtube.com/watch?v=JrkrwwoRtD

Ihttps://www.youtube.com/watch?v=oP-viCmMy7U

https://www.youtube.com/watch?v=iUQFzb997qs

so its about how we deal with different stream qualities etc. i have already made code in bitbucket but this recognises at match summary
page - and what i need is the recognition to happen in real time so once goals are scored it gives scorer and name of player. i had
another developer make a real time analysis but it doesn’t work well on lower stream quality like 360p. also I’m wondering if tesseract is
the best thing to use or if some other ocr library would be better, even if have to pay a license to use it.

So yeah thats my issue :) wondered if you had any time to let me know any guidance or even be involved in dev in solution, happy to pay some
money for your time :D

 thanks,
mark