The CVDataManagement project is a collection of Python scripts and utilities aimed at processing, analyzing, and managing image and video data for computer vision tasks. It provides functionalities like detecting duplicates or anomalies in images, processing video frames, and ensuring quality in datasets used for computer vision.
- Duplicate Detection: Detects duplicate images or near-duplicate frames in videos.
- Blurred Image Detection: Identifies and filters out images with poor focus.
- Exposure Detection: Highlights images with over- or underexposure.
- Frame Processing in Videos: Extracts and filters unique frames from video files.
- Utilities: Includes functions to draw text and rectangles on images for highlighting.
-
Command:
python keep_unique_frames_in_video.py --images_folder <input_folder_path> --output_folder <output_folder_path>
-
Arguments:
--images_folder: Path to the input folder containing videos or images.--output_folder: Path to the folder for output data.
-
Example:
python keep_unique_frames_in_video.py --images_folder "./videos" --output_folder "./output"
