Posted: 6/9/2017 6:09:17 PM EDT
|
This is a complete shot in the dark....GD is no place for this topic but oh well...
Any advice on improving the accuracy of Tesseract OCR for high resolution images taken from the desktop (1920x1080). I've tried rescaling, binary, contrast, etc. but the accuracy isn't where we want it. I tried a training app today that I found on the internet but wasn't happy with the results. Any advice on fixing Tesseract or another open source (non-cloud based) OCR solution that is great for computer text? |