Leginon Image Coordinate System
Added by Anchi Cheng over 14 years ago
Here is an e-mail from Thomas M.
we've now started doing some data collection using Leginon ... and observed that the very same acquired image (.mrc) is horizontally mirrored in the displays when viewed using the Leginon/ImageViewer and Ximdisp from the original MRC package. It seems that the coordinate system in Legionen starts at the top left corner (x goes down, and y to the right), whereas in Ximdisp the coordinate system is at the bottom left corner (x goes up, and y to the right). If I am not mistaken then such a mirrored image would cause a different handedness in the structures reconstructed. Maybe I am missing here something but would be great if you can let me know, whether we do have to mirror our images acquired in Legionon now ?
Replies (1)
Re: Leginon Image Coordinate System - Added by Anchi Cheng over 14 years ago
Hi, Thomas,
Relative to Ximdisp coordinate system, Leginon image display is indeed mirrored horizontally.
Leginon coordinate system is based on electronic reading and display device (like a TV) scanning order (left-to-right as the fast axis ; top-to-bottom as the slow axis), while Ximdisp is based on middle or high school math (left-to-right as the fast axis ; bottom-to-top as the slow axis).
If you had your camera set to orient such that an object at top right relative to the center in your viewing screen at the microscope shows up on the top-right of Leginon display, as we recommended, this image will need to be mirrored somewhere if you are used to Ximdisp coordinate system and want to use MRC software to process it as such. SPIDER, IMAGIC uses the same system as Leginon, EMAN's v2 and IMOD's 3dmod uses the one like Ximdisp.
Anchi