Deep-In-Sight/frame_extractor
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Repository files navigation
# Description This is a tool for selecting frames from a video. The pipeline consists of the following steps: 1. Extract frames from the video 2. Calculate Laplacian variance of the frames and select frames with high variance 3. Detect objects in the frames and save object masks 4. Detect vegetation and save vegetationmasks 5. Support static masks to apply to all frames (watermarking or fisheye contour) # Usage: ``` python3 main.py --help python3 main.py --video-names /path/to/video/*.mp4 ``` Output is written to /path/to/video/output