Welcome to !Trash!Trash (nɒt træʃ) aims to collect data on the amount of recyclable and compostable materials headed to landfills by equipping household trash cans with a camera on the underside of the lid to detect and categorize materials being discarded. |
A single recycled plastic bottle saves enough energy to run a 100-watt bulb for 4 hours.
The Components of !Trash
Image ClassificationRafael and Manasa focused on developing and training the image classification model used to identify household items. Rafael has also worked on incorporating openCV in order to use video feed to make predictions based on the classification model. Manasa also focused on interfacing the results with an arduino in order to give the consumer a visual signal.
|
BackendBrenda organize the data she received from the database of incoming information about whether the item is recyclable, compostable, or landfill into a spreadsheet. She used Python and SQlite and Google API.
|
FrontendMiles created and designed the website and embedded a fast-updating pie chart graphical representation modeling the data processed by the backend to display on the Global Analytics page.
|