Trying to get the bounding boxes of all recognized words using python-tesseract

[Mrinmoy Nath]

Hi,

I am trying to extract each word from a .png image (converted from pdf documents).

Using Python 2.7 and tesseract-3.05 APIs.

But for few of the documents instead of drawing the bounding box around a word Tesseract is drawing the same for a larger area and missing some of the words.

I am using 1111.png as input. Also find the output in 1111_op.png.

Could you please help me out to understand what could be the reason. 

Regards,

Mrinmoy

 

[Dhrumil Barot]

Did you solve this problem? I also have similar layout documents with handwritten digits.