These are chat archives for fiji/fiji

27th
Aug 2018
Varun Kapoor
@kapoorlab
Aug 27 2018 16:50
Hi, If i have a (t,X) image as RandomAccessIbleinterval<T> in pixel units (t = framenumber, X = pixel co-ordinate) and want to make a calibrated image along the (tcal,Xcal) so that tcal = sec and Xcal = micrometer, simple scaling does not seem to be correct way to transform the image. I should create a calibrated image along (tcal, Xcal) axes and then interpolate it to create a proper image. It seems like the kind of thing that must have been implemented in imglib2 somewhere, could you please point me where it is? thanks.
Jan Eglinger
@imagejan
Aug 27 2018 18:30
@kapoorlab setting a calibration for a spatial or time axis does not change the number of samples that you have on either of these axes. What do you want to achieve? An interpolation of “missing” values?
Varun Kapoor
@kapoorlab
Aug 27 2018 18:39
@imagejan Yeah, if I plot a Kymograph with rolled out (x,y) co-ordinates (say 500 pixels) on y axis and frame-number (say 200 frames) on x axis I get a plot of 200 by 500 but if I want to plot with correct time axis and say the time calibration if 1 frame = 5 seconds the image dimension would be 1000 by 500. If I do it that way I have missing values every 4 units along x-axis. Shouldn't that be interpolated?
Varun Kapoor
@kapoorlab
Aug 27 2018 19:49
if it has not to be interpolated then it is just a simple multiplication on the axes.