A python wrapper for using NoahApthorpe/ConvnetCellDetection from a windows machine. Due to incompatability between Docker for Windows and Virtualbox, ConvnetCellDetection is incompatible with Windows. There may be a work around using the older "docker tools", but docker tools does not allow mounting a virtual machine to a container in Windows, which made the process complicated.

Rather than set up the software on windows, this implementation sends data to a linux machine with ConvnetCellDetection installed, which gets around the incompatibility of the Convnet with Windows.

-install python 2.7

-clone the repository to a local directory

-setup a virtual environment with the packages listed in requirements.txt

change Hostnames, IP addresses, and passwords in ConnApp.py to match your linux server

run ConnApp.py from within virtualEnv

Open a file, it will get sent to the linux machine for cell detection. The machine will send back cells that the neural net found!

run visualizeTom.py from within virtualEnv to see outputs in a gui. You can then adjust roi thresholder and save the outputs. The output of these is compatible with imageJ.