/people-counter

Computer Vision Based People Counting Tool using Raspberry Pi and OpenCV

Primary LanguagePythonOtherNOASSERTION

Freeboard

My Jupyter Notebooks

My expanding set of Jupyter Notebooks.

Pull text from /home/jeff/blogging/content/ideas/.

What if you Jupyter environment isn't on your local computer, but instead on a remote compute accessible via TCP/IP? You want to open and manipulate an Jupyter Notebook running on the remote computer. This can be done by opening an SSH tunnel. This tunnel will forward the port used by the remotely running Jupyter Notebook server instance to a port on your local machine, where it can be accessed in a browser just like a locally running Jupyter Notebook instance.

On the remote machine, start the Jupyter Notebooks server:

# on the remote machine, start the jupyter notebooks server
jupyter notebook --no-browser --port=8889

On the local machine, start an SSH tunnel:

# on the local machine, start an SSH tunnel
# run in background: ssh -f -N -L localhost:8888:localhost:8889 remote_user@remote_host
# run in foreground: ssh -N -L localhost:8888:localhost:8889 remote_user@remote_host
ssh -N pi@BlueRPi -L localhost:8888:localhost:8889

Now enter localhost:8888 in your favorite browser to use the remote Jupyter Notebook!

Within Chromebook ....

  1. In one window, login to desktop -- cd Jupyter-Notebooks ; jupyter notebook --no-browser --port=8889
  2. In 2nd window -- ssh -N jeff@desktop -L localhost:8888:localhost:8889
  3. In 3rd window -- gnome-www-browser
  4. Now enter localhost:8888 in the browser and now you can access the remote Jupyter Notebook!

Sources

The people counting algorithm

Fast video processing

Loading OpenCV