Python wrapper for Engine Cube + Getting font size using Tesseract Engine (in Python)

[aliaksei...@epam.com]

I've heard that Engine Cube uses better model than original Tesseract Engine. So is there any python wrapper for Engine Cube?

Also is there any way to get font size using Tesseract Engine? I come to the idea of using hocr mode and get font size of every word from generated html, but maybe there is a better way to do that?

[rohit saluja]

I am also searching for training the tesseract for cube mode. Just posting so that I get to know if anyone replies here.

Also where can we get knowledge for python wrapper for tesseract?

[Tom Morris]

On Monday, June 13, 2016 at 4:11:51 PM UTC-4, aliaksei...@epam.com wrote:

I've heard that Engine Cube uses better model than original Tesseract Engine. So is there any python wrapper for Engine Cube?

The recognition engine is an internal implementation detail, so you should be able to use the same API and Python wrapper. As an aside, I'm not sure it's true the cube recognizer is considered universally "better."

 

Also is there any way to get font size using Tesseract Engine? I come to the idea of using hocr mode and get font size of every word from generated html, but maybe there is a better way to do that?

The font information, including size, is available through the API.  You can see how it's fetched for hOCR output here:

https://github.com/tesseract-ocr/tesseract/blob/master/api/baseapi.cpp#L1552

On Tuesday, June 21, 2016 at 3:09:35 AM UTC-4, rohit saluja wrote:

I am also searching for training the tesseract for cube mode. Just posting so that I get to know if anyone replies here.

I don't think training tools were ever released for cube mode.

 

Also where can we get knowledge for python wrapper for tesseract?

The top three search hits on Google all look relevant. #1 is a package on PyPI the standard Python package index.

Tom