Inspiration

One day, my fiancee ask me if there's a little sensor that could be attach to an item to detect its location because her coworkers sometimes forgot to put the items back to where they took them, and that makes their work very inconvenient. Therefore, I came up an idea that to use cameras to detect objects.

What it does

This system is using cameras(No matter security camera or street camera) to detect objects' categories, sizes, colors and locations and store them onto the database in order to find them easily by typing or saying some key-words or descriptions(Category, color, shape, size... etc.).

How I built it

To do so, we use Python with Tensorflow(Machine Intelligence Platform) to build up the system under Linux for recognizing objects and calculate their sizes, shapes, colors and locations.

Challenges I ran into

It was very difficult to set up the system because I am not good at Linux, I only know some basic Python codes by self-learning, and the module we were trying to install had a lot of problems to solve.

Accomplishments that I'm proud of

The idea interests people especially the mentor.

What I learned

I improved my Python, troubleshoot skills, critical thinking and get to know Tensorflow

What's next for POC(PieceOfCake)

Build up this system, train the system to detect a wider variety of objects more precisely.

Built With

Share this project:

Updates