Saturday 04/26/14

coding

• working on rewrite of chromewarp code: ComputeChromeWarp.m
• this is now a properly functioning and tested function. Needs to be integrated into STORMfinder.
• should operate with a simplier menu — would like to be able to specify a list of alternate parameters files as well as have defaults.
• this code should run just based on what ir / vis datasets it detects without the user having to specify the channels. I think in practice this will work better than the arbitrarily generalized N movies of M combinations of channels.
• meanwhile the code is also faster due to the removal of splitdax activity. And the code is shorter and hopefully easier to read.
• some gentle generalization might add more header in the near (not so near?) future.
• C08 data set chromewarp is very difficult, way out of line. Possibly raw image based autocorrelation would help. But it’s really not that necessary I think, maybe we can just avoid the beads manually? let’s try and see.
• Did get improved localization on previously analyzed regular quadview IR/Vis dataset using my improved MatchFeducials.m function (still point-based).
• updated matlab-functions Warp2BestPair to use the revised/improved function MatchFeducials.