Hey everyone!
I moved this from the Gamma forum
As my long term goal is to have this be an all-round solution for anyone working on integrating OCR into their EliteDangerous app, it's spun off into it's own project!
https://github.com/CapCap/EliteBrainerous
Over 99.99% accurate
This isn't a place to discuss whether or not such a thing should exist- I wish I could delete posts like that because it's needless clutter, and there are a lot of other threads that exist for that kind of debate.
This is a place to ask questions about the OCR, get implementation help, report bugs with it, etc. I hope this becomes a resource for fellow devs trying to do OCR in EliteDangerous
If you're looking for a sexy, sleek project using EliteBrainerous for parsing market data, look no further than Seebek's EliteOCR: https://forums.frontier.co.uk/showthread.php?t=68771
(sample output from the slightly less accurate RBM- I recommend using the logistic classifier)
ToDo:
1. Expanding EB to classify characters in addition to digits (in progress!)
2. Improve the character segmentor (don't really want to do this myself but will if I have to!)
3. Implement own NN backed sliding window text-detector to determine own text-region bounding boxes (same as #2!)
4. Extend this to actually capture the screenshots internally, recognize the screen, and provide an API to whatever data is contained therein (i.e if(eb.currentScreen() == 'market'){ print eb.parseCommodityData() })
PS: Hey Frontier! I'm available for hire or freelance ;-)
I moved this from the Gamma forum
As my long term goal is to have this be an all-round solution for anyone working on integrating OCR into their EliteDangerous app, it's spun off into it's own project!
https://github.com/CapCap/EliteBrainerous
Over 99.99% accurate
This isn't a place to discuss whether or not such a thing should exist- I wish I could delete posts like that because it's needless clutter, and there are a lot of other threads that exist for that kind of debate.
This is a place to ask questions about the OCR, get implementation help, report bugs with it, etc. I hope this becomes a resource for fellow devs trying to do OCR in EliteDangerous
If you're looking for a sexy, sleek project using EliteBrainerous for parsing market data, look no further than Seebek's EliteOCR: https://forums.frontier.co.uk/showthread.php?t=68771
(sample output from the slightly less accurate RBM- I recommend using the logistic classifier)
ToDo:
1. Expanding EB to classify characters in addition to digits (in progress!)
2. Improve the character segmentor (don't really want to do this myself but will if I have to!)
3. Implement own NN backed sliding window text-detector to determine own text-region bounding boxes (same as #2!)
4. Extend this to actually capture the screenshots internally, recognize the screen, and provide an API to whatever data is contained therein (i.e if(eb.currentScreen() == 'market'){ print eb.parseCommodityData() })
PS: Hey Frontier! I'm available for hire or freelance ;-)
