This repository contains code to extract Chinese hard subs from the TV series 他来了请闭眼 (Love Me If You Dare). For further information please see this post on my blog.
To get started, install OpenCV, Tesseract, the chi_sim data pack for Tesseract, and PyOCR. The following commands will work on Arch Linux:
sudo pacman -S opencv python-numpy tesseract tesseract-data-chi_sim
sudo pip install pyocr
Then try running ./main.py --test-all to test the extraction algorithm on all test cases. To run it on a video file, you'll need to track down a 1280x720 video of one of the 他来了请闭眼 episodes with white hard subs at the bottom, similar to the test frames.