0

My dataset is annotated using rotated bounding box annotation tool having (x1y1 x2y2 x3y3 x4y4 class category) is label format (same as DOTA).

My data directory contains:

  1. images
  2. train (txt labels)
  3. val (txt labels)
  4. test (txt labels)

I want to augment training folder along with annotations. What is the best tool/library that can help me achieve this?

user181925
  • 3
  • 2

1 Answers1

0

you can use the imgaug library in Python.it is powerful library for image augmentation, including support for bounding box transformations. documentation:https://imgaug.readthedocs.io/en/latest/

Amith A G
  • 378
  • 13
  • Thanks for your suggestion, I am wonder it will work with DOTA like set or not? I am sure it is good for normal bbox like xywh but in my case it is xyxyxyxyclasscategory. – user181925 Jul 11 '23 at 08:17