Subtitle lines with larger Levenshtein ratios than this threshold will be merged together. Make it closer to 0 if you get too few words in each line, or make it closer to 100 if there are too many excess words in each line. The default value 65 is fine for most cases. Words with lower confidence than this value will be discarded. You can read more about Tesseract language data files on their wiki page.Ĭonfidence threshold for word predictions.

Language files will be automatically downloaded to your ~/tessdata. lang='hin+eng' for Hindi and English together.

Note that you can use more than one language, e.g. 'HanS' for simplified Chinese) are supported. 'eng' for English) and all script names in this repository (e.g. You can extract subtitles in almost any language. Write subtitles to file_path save_subtitles_to_file ( video_path : str, file_path = 'subtitle.srt', lang = 'eng', time_start = '0:00', time_end = '', conf_threshold = 65, sim_threshold = 90, use_fullframe = False ) Return subtitle string in SRT format get_subtitles ( video_path : str, lang = 'eng', time_start = '0:00', time_end = '', conf_threshold = 65, sim_threshold = 90, use_fullframe = False ) Install Tesseract and make sure it is in your $PATH It takes 3 minutes on my dual-core laptop to extract a 20 seconds video. print ( get_subtitles ( 'video.mp4', lang = 'chi_sim+eng', sim_threshold = 70, conf_threshold = 65 )) # example.py from videocr import get_subtitles if _name_ = '_main_' : # This check is mandatory for Windows. Extract hardcoded (burned-in) subtitles from videos using the Tesseract OCR engine with Python.

Youtube video subtitle extractor