This is not really an issue.
We are using SyMBac for producing training data for fluo imaging/segmentation with Delta. It does seem that Delta is getting confused when all cells in one trench always have the same fluorescence (I have no hard evidence for this). However, I included this in my SyMBac version. I added
self.fluo_variation = np.random.uniform(0.5, 1.5)
to the constructor of the cell class. Then I added this to the timeseries properties generated in gen_cell_props_for_draw. Finally, in draw_scene, I did
OPL_cell = raster_cell(length=length, width=width, separation=separation, pinching=pinching) * fluo_variation
I am sure there are smarter ways to do this, but it did the trick. Happy to add this to the newest version at some point.
This is not really an issue.
We are using SyMBac for producing training data for fluo imaging/segmentation with Delta. It does seem that Delta is getting confused when all cells in one trench always have the same fluorescence (I have no hard evidence for this). However, I included this in my SyMBac version. I added
self.fluo_variation = np.random.uniform(0.5, 1.5)to the constructor of the cell class. Then I added this to the timeseries properties generated in
gen_cell_props_for_draw. Finally, indraw_scene, I didOPL_cell = raster_cell(length=length, width=width, separation=separation, pinching=pinching) * fluo_variationI am sure there are smarter ways to do this, but it did the trick. Happy to add this to the newest version at some point.