1) Use python to programmatically download/read a 16-bit TIF micrograph with fluorescently labelled cell nuclei from the following location:
https://github.com/Shutterbug2109/Microscopic-Image-analysis/raw/main/Sample.tif Once loaded display the images, scale them appropriately to ensure the cell nuclei are visible. NB: Make sure to import all the necessary libraries allowing you to load raw TIFs.
2) Use an arbitrary threshold value (defined manually) to convert nuclei from the grayscale image to a binary mask. Once done - display the mask and the raw image side-by-side.
Optional: find optimal threshold value automatically.
3) Measure and display the pixel area of each nucleus. Avoid thresholding artifacts, which differ significantly from the nuclei in certain aspects.
4) Using Numpy, create a Python generator producing a 96 by 96 pixel-sized random crop from the original grayscale image.
Optional: Create a generator compatible with the built-in Keras, Tensorflow or PyTorch classes.