Skip to content

Latest commit

 

History

History
41 lines (24 loc) · 957 Bytes

README.md

File metadata and controls

41 lines (24 loc) · 957 Bytes

labelTool

This is a labeling tool to label objects to train neural networks.

Requirements

The project requires

  • python3
  • tkinter
  • pillow
  • xml
  • opencv
  • argparse

Run tool

To run the tool. Use the run.sh script to run the samples.

or

python3 boundingBox.py --input_directory=/path/to/images --output_directory=/path/to/output --classes=/path/to/class_list.txt --choice=3

set --choice to 1 for YOLO darknet format

class x y w h

set --choice to 2 for imagenet format without XML

number of bounding boxes x1 y1 x2 y2

set --choice to 3 for Pascal VOC format with XML output

Do not press back when using --choice 1 and 3 as they are not converted back to x y coordinates to display and hence will delete the labels

The --classes parameter needs a text file with the class names. It picks up the class name to add to the XML.

Acknowledgements

The base code if from https://github.com/puzzledqs/BBox-Label-Tool