Hi,
I've been using TrackMate for z-stack timelapse images of variable sized
fluorescent 'blobs.' There is no single correct blob diameter but all the
algorithms TrackMate uses (LoG, etc) require one which leads to unnecessary
splits. Does anyone have any advice on how to fix/get around this?
Possible routes I've looked into: (none are ideal)
* Modify the LoG blob detection to give "multi-scale blob detector with
automatic scale selection" as detailed in the wikipedia page
<
https://en.wikipedia.org/wiki/Blob_detection#The_Laplacian_of_Gaussian>
(bottom
of linked section)
(Any advice on this would be appreciated. I am daunted but working on
extending imglib2 to allow for higher dimensional optimization of
localextrema, etc)
* Additional FIJI plugins which allow manual modification and refinement
* I tried MTrackJ but it doesn't allow splits
* I tried TrackMate's semiautomatic tracking but had difficulty with the
interface and it's less exact tracking algorithm
Thanks for reading this, I appreciate it!
Jonah
--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html
Locked
