center of rotation finder#
Behavior#
This widget is used to compute the center of rotation for a set of acquisitions.
You can choose to use projections or generate a sinogram from them to compute it.
The axis widget behaves as a pop-up if it requires some information to process. This will happen until you ‘lock’ the axis with a fixed value.
Note
Tomwer displays the value of the center of rotation (COR) in \([- \frac{framewidth}{2}, \frac{framewidth}{2}]\). However, Nabu uses the COR value in \([0, framewidth]\). This is handled internally, but if you want to pass this value to Nabu or pyshst (or check output files generated by Tomwer), you should add \(\frac{framewidth}{2}\) to this value.
Center of rotation from radiographs#
The interface for computing the center of rotation from radiographs is as follows:
The options for center of rotation calculation are divided into three parts:
calculation: Used to define the algorithm and mandatory parameters (like near position) for the near algorithm.
You can find a short description of the algorithm in Algorithms.
options: Used to define some advanced controls for the algorithm. The available options depend on the selected algorithm.
input: Used to define the projections for the center of rotation calculation. The ‘0-180’ and ‘90-270’ angle modes attempt to select the closest projections to 0 and 180 degrees or 90 and 270 degrees for calculation. The third option allows manual selection of projections.
Center of rotation from sinograms#
The interface for computing the center of rotation from sinograms is as follows:
For sinograms, you should first select the type (emission or transmission). If transmission is selected, the algorithm will compute log values.
Then you can pick a radiograph line, and when you select ‘compute’, it will: * Extract the sinogram for the selected line. * Compute the center of rotation for this line.
You can also select the subsampling value. As datasets can be large, you can choose to take only one of every ‘subsampling’ frames.