Today, I discovered a useful tool that will delight fans of TV series, anime, and subtitled movies. Many of us have experienced the frustration of not being able to retrieve the text of the subtitles embedded in a video, making it impossible to correct mistakes or automatically translate the subtitles into another language.
Thankfully, with the program called VideoSubFinder, you can extract the text embedded in a video and use it to create new subtitles or translate them.
The process is simple: the program automatically detects frames with embedded text (also known as “hardsub”) and records the positions and timings. Then, it generates images of the text with the background removed, which can be used with an optical character recognition (OCR) program like FineReader or Subtitle Edit to create full subtitles with the original text and timings.
This program is compatible with Windows and Linux operating systems, but it requires certain dependencies to be installed to ensure proper functioning. For Windows users, it is necessary to download the Microsoft Visual C++ Redistributable 2022 library, while Ubuntu 20.04.5 LTS users must install libgtk-3-0.
This application is not only useful for those who deal with subtitles but also for individuals working in the field of translation or for those who simply want to extract text from videos. It is an incredibly handy tool to have.