king of trash-bin

Trash recycling is posing threats to many countries. We combine image labeling and trash-bin management system to monitor status of trash-bins and classify trashes efficiently.

Comment

Inspiration

We noticed that trash in trash-bin are not well classified and this may lead to great waste of resources. Meanwhile, it’s typically hard to force people to classify trashes correctly. Monitoring status of trash-bins in real time is tedious for cleaners. Thus we suggest a better way to manage status of trash-bins based on intelligent classification of trashes.

What it does

  1. Monitor status (fill/empty/half-filled) of trash-bins and send notifications to cleaners when it gets filled.
  2. Classify the trashes into corresponding category without human intervention

Accomplishments that we're proud of

  1. Build a simulation program to simulate mechanical architecture of new-generation trash-bin
  2. Collaboration together to make IOS app within one day

What we learned

  1. Realistic situation is much more complicated than we expected. You should carefully consider all possible cases.
  2. We may save much time and efforts with knowledge of some existing libraries or packages.
  3. We get access to various techniques like Firebase, Google Vision, AlwaysAI.

What's next for king of trash-bin

  1. We will apply it to wider areas, instead of USC campus.
  2. We will try training machine learning model to improve accuracy of category prediction
  3. Design cooler trash-bin(adjust the capacity for different types of trash)

Built With

Share this project:

Updates