You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I suggest building the registration module in the following iterations.
Extend the deskewing CLI (after merging Generalized deskewing for mantis #6) to compute the rigid transformation in XY and XZ planes between label-free and one fluorescence channel to get rough alignment.
Refine the alignment further with phase correlation or other algorithms as needed.
In this case, it may make sense to register the volumes via writing the Coordination Transformation metadata. @edyoshikun can you link the issue/code snippet on iohub on how to do this?
We also needs to figure out how to collect the registration data. As LF and LS arms can be focused independently, we'll likely need to collect registration data every time, and allow for Z drift over time. Image-based autofocus that adjusts O3 in each arm may help here.
I suggest building the registration module in the following iterations.
In this case, it may make sense to register the volumes via writing the Coordination Transformation metadata. @edyoshikun can you link the issue/code snippet on iohub on how to do this?
Looks like @talonchandler can implement (1), and @edyoshikun can review.
The text was updated successfully, but these errors were encountered: