The software is intended to count pixels in multiple z-stacks of images. It is the companion analysis software for the high-content assay described in Murali et al., BMC Cancer, 2019, previously made available as a preprint.
Usage:
python process_images.py <FOLDER> <OUTPUT_FOLDER> <CHN1> <CHN2>...
where CHN1, CHN2, etc are the names of the channels.
Example:
python process_images.py DMSO output Nuc Apop EtHd
Inside the folder, the images should be organized as:
xy<XY_COORD>z<Z_COORD>c<CHANNEL_ID>.tif
Example:
xy01z001c1.tif
When running the usage example with the example image names, c1 gets mapped to the name Nuc, c2 gets mapped to Apop, and c3 to EtHd. The result files will be formatted appropriately.