Closed bruno-31 closed 1 year ago
Hi, yes that's true, that's how it is implemented for now. That's because I mostly used TV for CT reconstruction for which images are typically square, see https://github.com/eboigne/PyRAMID-CT.
I don't really plan to update the code for non-square images right now. But, as a quick work-around, I'd recommend filling in your images with zeros into square images. You'll loose some performance, but that should be reasonable if your images are close to square.
Thank you !
Hi, thank you so much for the nice work! It seems that it is only possible to use square images in your code, why?