0

I have some problem training segmentation models.

I have some UAV images(in tif format), which I cut them into pieces like this:

enter image description here

And I made some mask in ENVI like this(still tif format):

enter image description here

When I transform it to jpg format, It's like this:

enter image description here

Enough introduction here, and my problems are:

  1. If possible, I want to use Tif format images to train the model(like U-net FPN...), is there any exiting library support tif as a input?

  2. You can see my images are slanting, part of them are transparent. Is it legal to input training images like this?

    I will be more than thankful if you can offer me some places to find examples alike. Thanks!

I cannot use cv2.resize to change the input size of tif mask. And I have some problems loading images data. Anyway my original results are awful, so I need some examples and tutoral.

Viktorrrr
  • 13
  • 1
  • all the DL libraries offer to let you handle the loading. `imageio` can handle TIFF, a few other libraries can too (OpenCV, PIL, ...). if your network takes 4-channel input, go ahead. if not... think about what you want the transparent background to be, because it _is not_ going to be transparent. it might be presented as black. -- a [mre] would help others help you. it also helps you sort your thoughts and debug your problem. – Christoph Rackwitz Feb 06 '23 at 10:20
  • Thanks for your time and kind suggestions, Christoph! I will try to read that ,and make my problems easier to understand. – Viktorrrr Feb 06 '23 at 17:42

0 Answers0