Leginon

h1. Black Stripe Detection

For users of the K2 camera. At NYSBC we occasionally lose a channel during collection, resulting in a black stripe in the image. More rarely, counted movies start coming out the wrong size, off by 2 pixels in the X and Y axes. This node will detect both types of problems and throw an warning message. Particularly useful feature if slack messaging is in use.

h2. Adding the node

Binding needs to be made from Exposure to the Black Stripes Detector Mode. Binding is AcquistionImagePublishEvent.

h2. Use

Control panel has 2 options, whether to detect black stripes and whether to pause on error. Turning it on means every exposure image will be analyzed for lack of signal in all 8 stripes. Will output the mean and sd of each stripe in the text window. A black stripe is defined as having mean=0 and sd=0. Additionally tests the size to make sure it is K2 standard size (3710 x 3838)c

h2. Testing

Works on images in both orientations (portrait and landscape). Works on super-res images (bins to counting before testing). Does not seem to affect collection rates. Calculations are relatively simple and no pausing is needed during calculations

h2. Notes

Note: Pause feature is not yet implemented in the code. Can be checked but no pausing is done. May add a feature to only check every N images.