The team consists of Nick, Andrik, Ian and Thanh. Our project idea was inspired by recent changes to what can be recycled in a home. Over the course of HackED we are learning how to use Tensorflow to process an image captured on a USB webcam by openCV and then communicate with an Arduino on the sorting machine to place each item in the correct area. The primary challenge has been learning how to effectively process an image and automatically determine whether it is recycle, compost, or garbage with significant accuracy.
HackED-2020-RecycleProject-NickAndrikIanThanh
An automated system implementing a neural network with image recognition to recognize and then sort recycling, compost, and garbage.
Updates
10:30am, robot can accurately place an item in the correct location once a signal has been received from the neural network. Neural network can recognize certain items very well. In order to increase the accuracy of the system we would need to create a much larger data set (currently have around 7000 images) as well as have much more computing power to train the neural network with the larger data set.
Leave feedback in the comments!
Log in or sign up for Devpost to join the conversation.
Log in or sign up for Devpost to join the conversation.