Thanks for the quick answer. I would be very happy to have this transformation field.

On one of my channels, I have an anti-betagalactosidase staining against lacZ that allows me to see hit nuclei in Gal4 lines. I want to detect them and process them in the reference brain.

My idea is to detect these dots on the original image and not on the warped image, because the image quality is worse (clusters of nuclei look already blurry in some places in the original image). Then I would like to compute the new locations and sizes of the nuclei, based on the locations on the original image and the DF.

So I just need the DF as output, not as input. I will apply the DF to a kind of binary datas (the list of detected dots)

The DF only for the warping transformation should be enough. The affine transformation does not change the nuclei so much. I could run my nuclei detector on the global aligned images.

If it takes too much time to update the whole site, don't hesitate to email me an updated version.

Thank you very much.

Thomas