ACE(All S(C)eeing Eye)

Situation descriptor using Google Cloud Vision API and Rasberry Pi which doubles as a home automation system.

Comment

Inspiration

The inspiration was that we wanted to follow the mantra of more is less where we wanted to make more from just a raspberry pi.

What it does

It recognises frames from the pi camera and recongizes the objects in them, filtering them on an on-person basis as well. This has a variety of different utilities

How we built it

We used openly available opencv libraries and google cloud vision api

Challenges we ran into

OpenCV modules were not present always. PIp upgrades ran into issues. Vectorized packages didnt appear as fast when run in raspberry.

Accomplishments that we're proud of

ACcurately predicted a host of different objects with relatively less turnaround time considering the computation performed in raspberry. We should be able to improve the model with a more powerful or a substanicially hardwired

What we learned

Interfacing with raspberry pi is not as easy as I thought.

What's next for ACE(All S(C)eeing Eye)

We can improve the model and focus more on the home automation aspect of ACE by looking into automated refrigerator restock, automatic temperature control and automatic light control.

Built With

Share this project:

Updates