Text2image Issues for Tamil Unicode Font Sundaram - 0807

[Mugunthan]

I've been making tif/box files for Tamil character recognition using text2image on windows. I came across some issues for this Unicode Font Sundaram-0807 with text2image. 

Issue1: *Some characters in the tif file doesn't match with the text file. 

Issue2: Some characters in the tif file match with text file only at some instances.

I tried changing the size of generated tif file but it doesn't help. Please see the attached screenshots and the files for the font Sundaram-0807 - Size 12

Ps: I've used other tamil unicode fonts such as Latha, Akshar and TheeneeUni, they all worked perfectly.

[shree]

>Ps: I've used other tamil unicode fonts such as Latha, Akshar and TheeneeUni, they all worked perfectly.

As you recognize in statement above, problem is with the font not tesseract.

[Mugunthan]

Hi Shree,

Thanks for your reply. I've just tried this procedure using the same font in the Linux environment and it worked fine. I believe there should be some issue with Windows version text2image. I'm using ub mannheim version as suggested in github.