Like Andrea says, use masks to exclude areas of the image that have no value, or are causing alignment issues.
We have recently updated our Zephyr Foundation course and included a lot more content that covers using masks and we teach an example where we move the subject in the scene to shoot both sides, and use masks to force Zephyr to ignore the background.
So we can use source images like these:
And this:
And derive a model like this:
This bulls skull dataset is included in the course as a download, along with some sample masks created in Photoshop along with the Masquerade techniques taught.
Leave a comment: