Search code examples
tesseractambiguitytraining-dataunichar

Tesseract ambiguity files work different for editing

I want to edit some text like Female and Male because When I test them I found them as FemaIe and MaIe (I mean with Capital I not small L (l) ). And I want to solve this issue using ambfile like;

v1

6_tab_F_e_m_a_I_e_tab_6_tab_F_e_m_a_l_e_tab_1

4_tab_M_a_I_e_tab_4_tab_M_a_l_e_tab_1

But When I retest my results , they were worse. I found the Female as F and Male as M. What I am doing wrong ? To use amb file like that is a wrong idea?

Solution

  • The fields should be tab separated, according to Tesseract Training Wiki.