User Guide for Local Video Hard Subtitle Extraction Tool
In video processing, using speech recognition to automatically generate subtitles (such as with the Whisper model) is now very convenient. However, if subtitles are already "burned" into the video frames (hard subtitles), extracting them is not easy. Using online OCR interfaces often leads to call limits or errors, while paid services can be costly.
To solve this problem, this tool integrates two excellent open-source projects to create a fully local, completely offline video hard subtitle extractor. It can directly recognize text from video frames and automatically generate .srt format subtitle files. Currently, it supports Chinese, English, Japanese, and Korean languages.
Download link: Link
Must be unzipped in an English folder; a Chinese directory may cause errors.
Core Principle Breakdown:
- VideoSubFinder: Responsible for quickly locating frames in a video where hard subtitles appear and capturing screenshots.
- PaddleOCR-json (Based on Baidu PaddleOCR): Responsible for recognizing text from these screenshots and converting it into text.
Key Features
- Visual Region Selection: After importing a video, you can directly use your mouse to drag and select the subtitle area in the preview window, avoiding interference from other on-screen text (such as logos or watermarks).
- CUDA GPU Acceleration: Improves the efficiency of text recognition.
- Multi-language Recognition: Supports hard subtitle extraction for Chinese, English, Japanese, and Korean.
- Operating Environment: Currently only supports Windows 10 and Windows 11 systems.
How to Use
Step 1: Launch the Software
Download and unzip the software package. In the extracted folder, double-click ocrsp.exe to run it. 
Step 2: Import Video
On the left side of the software interface, click to select the video file you want to process (make sure the video contains the hard subtitles you need to extract).
Step 3: Select Subtitle Area
- In the video preview area on the right, hold down the left mouse button and drag to select the subtitle area.
- To ensure the selection is accurate, you can use the "Step" button below the preview window to fine-tune the video frame. Switch to other time points to confirm that the subtitles are fully included within the box at different positions.
Step 4: Start Extraction
Once confirmed, click the "Start" button. The log window at the bottom of the software will display the processing progress in real time. Just wait for it to finish.
Step 5: Get Subtitles
After extraction is complete, the generated subtitle file will be automatically saved in the same folder as the video, named [Original Video Name]-ocr.srt.
Acknowledgments
This tool would not be possible without the support of the following open-source projects. Special thanks to: