RPI:
The folder imgserve should be on your rpi - make sure your camera is enabled and mosquitto -d is used to init the mqtt server.
Add the correct IP addesses to run_analysis.py then run it (use & for background run, or nohop on ssh)
Computer:
mosquitto -d
python classifier.py
You can also run 'mse_graph' while its running for a live graph of mse & err vs count
Make sure your model is trained, I am not uploading the training data because its large, but the output_graph is there from my car classifier
#TODO
- Save all frames from motion detection?