OCR

PTMuxer provides optional OCR of sup files.

Currently this can be used to ID tracks based on the content.

Enabling requires configuring at least one of two engines.

Tesseract OCR

Tesseract OCR Provides the highest level of language compatibility in PTMuxer.

A downside is that it requires packages on the host system. Meaning that typically root would access would be required on a Linux Box.

Requirements

Visit repo for instrucions on how to install required binaries

https://github.com/sirfz/tesserocr

However this package does require that tesseract binary be installed on the system

If you don't have sudo or dont want to install the requirements. Then easyocr might be a good choice

Install

You can install the module with

CPPFLAGS=-I/usr/local/include pip3 install tesserocr

Windows Support

Windows Releases with all required libraries are provided.

They are specific to each python version

https://github.com/simonflueckiger/tesserocr-windows_build/releases

EasyOCR

As the name implies this program is easy to install, and use

It does not provide nearly as many languages as Tesseract. However, it still supports 80+ languages. So you will find that the most popular languages are included.

Windows Support

You may get errors mentioning torch on windows machines

If this happens manually configuration of the torch module may be required

Running this command may fix the issue

pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu116