ashnair1 / dotadevkit Goto Github PK

i tried to split images with scale=0.5 but it didnt work and used scale=1
thanks for you help in advance

in the dotadevkit, don't find ResultMerge_multi_process

Hello,
How to split the DOTA test set images that do not have labelTxt files?
Thank you

dotadevkit does not currently support visualisation via the CLI. Need to support this.

Ref: https://captain-whu.github.io/DOTA/dataset.html

I have a big problem and already tried many ways to find a solution for this error, but it still remains in my codes.
I am trying to reimplement the " BBAVectors-Oriented-Object-Detection" and firstly used "DOTA_devkit" but always facing
"import _polyiou ModuleNotFoundError: No module named ‘_polyiou‘"
Then I switched to "dotadevkit" but still I have the same issue. The "dotadevkit" works fine for the other tasks.

I use Python 3.6.6 on Ubuntu.

I wish to receive a suggestion to solve this problem.

Ive installed dotadevkit using anaconda on a windows machine.
Any idea how I go about running it, Ive tried running it on the CMD/anaconda prompt, it doesnt work.

How do i get it up and running. I wanna split the images and their respective annotations

dota_utils.py currently contains some convenience functions for handling filenames and converting box coordinates to different formats.

[] Remove file functions. pathlib is a much better alternative.
[] Refactor other functions. parse_dota_poly functions are the only ones actively being used for the core dotadevkit functions.

How can I have an annotation in background class (like COCO) with DOTA dataset?