This repository contains a python script that can convert JSON format of image annotations to png images. VIA stands for VGG Image Annotator.
Link: https://www.robots.ox.ac.uk/~vgg/software/via/via.html
- Select Images you want to annotate.
- Upload them.
- Define shapes such as Circle, Polygon, Polyline, Point.
- Now annotate each image and assign them a shape.
- Assign only one shape to one annotation.
- Once all images are done annotating, save the project with appropiate name as JSON file.
- Now run the generator.py with necessary file paths.