After not bothering with optical charactor recognition for a long while, I decided to give it another shot. An image scanned at 600 dpi and saved as a .png actually translated very well to text. Lower case "i" showed an accent tick instead of the dot about 50% of the time and about 10% of the time, lower case "L" was converted to upper case "i". Superscript is even more problematic with the exponant "-??" showing up as "-`" or "-1", but never correct.
Can anyone point me towards some documentation that might help me decrease errors? -- Regards, Ernie
