I regularly record data at 30 Hz for one hour at a time (over 100,000 frames). I am recording an object that moves around minimally (less than a hundred pixels). Each frame is stored as a .tiff file.
I'm trying to efficiently generate a "stabilization matrix." For example, if the image has shifted two pixels to the left and one pixel up, then I want to generate a matrix with a row as:
[2, 1]
to signify that image needs to be shifted 2 units on the x-axis, and one unit on the y-axis. Each row in the matrix would represent the necessary "shift" in this way.
Is this possible? I am open to using any language or platform. I also have access to a cluster at my university. I've been passed down a code that's written in Matlab, but it takes about 12 hours to run. I'm hoping to find a more efficient solution. Any pointers in the right direction would be greatly appreciated.
Thanks in advance.