Ask Your Question

Revision history [back]

Filtering a text image for pytesseract

Hello there,

Like the title says I'm having issues with moving on with my project as I'm trying to filter out text and numbers (numbers in specific) from an image captures by a raspberry pi so I manages to get through all the setup of the pillow, openCV, pytesseract, and picamera libraries (but still partly as I have to edit the code in the IDE and run it from the terminal ("python file.py") as I still don't know how to setup the path for the IDE's to search for the modules)

But after capturing the image and filtering it a little with tresh hold value and grayscale I can't manage the tesseract to realy read the image and convert it to text....

Here are some of the progress made:

The unfiltered captures image:

image description

after the filtering:

image description

BTW I'm using a raspberry pi zero W with running debian.

May I have any hints to how to move on with this?