Using the same channel mask for multiple images

[zakiindra]

First of all, I want to appreciate this work. It is very useful, and I'm currently using it for my research. I want to use the same cloud pattern for multiple images, and I believe it is returned as cmask in the example. Is it possible to use the same cmask for different images? I am evaluating performance of ML model robustness against cloud and I want to apply the same cloud pattern to many images to observe if different patterns influence the performance. Is there any way I can do it?

Thank you.

[mikonvergence]

Hi @zakiindra, thank you!

Yes, it is possible and I have done this before for my research I think.

Please have a look here: https://github.com/strath-ai/SatelliteCloudGenerator/blob/d0787cbdc1b39a7976f89aa085f1430f3207eda6/src/CloudSimulator.py#L328-L333

• You can get the cloud component by setting return_cloud to True
• You can then use the mix() method as shown above to mix the returned cloud with any image
• Ultimately, the current code is not best adapted to it, but it is definitely possible by doing this as described above - if you have some ideas on how the current package could be expanded to allow this more easily, feel free to propose!

Let me know if this answers your doubts.