1. For tesseract 3.02, after installation I see there's a pre-trained eng.traineddata file in the tessdata folder. How is this file generated? What font does it target? Can I blindly use it for my OCR application?
2. For tesseract 3.03, I see there's a new option "--print-parameters" for the tesseract executable. There're more than 600 parameters. How am I supposed to use them? If I need to tune them, how?
3. During my experimentation, I see tesseract works better for some font type than other font type. Is this true? Which font has the best precision?