/vcr-labeling-tool

Labeling tool for video content recogntion

Primary LanguagePythonMIT LicenseMIT

Labeling Tool for VCR

Labeling tool for video content recognition based on deep learning

This provides a convenient GUI for non-tech users to draw bounding boxes around targets in the image.

Functions

  1. Load image folder for labeling
  2. Select content type of labeling
  3. Adjust size of the bounding boxes
  4. Add new bounding boxes with label from each .names file
  5. Delete unnecessary bounding boxes
  6. Read/write image information from the json file

Environment setup

Installation

  • Anaconda latest version
    • python 3.6 or above
  • Python packages
    • pip install -r requirements.txt

Creating an executable file for Windows OS

  • Pyinstaller latest version
    • pyinstaller -w -F vcr_labeling.py