An image classifying model segregator.ipynb will segregate the garbage into bio-degradable and non-biodegradable waste . Using RaspberryPi and Servo Motor , we implemented the model by building a small dustbin , the top plate of which can flip +90 or -90 degrees depending on the category of waste .
Dataset we used contains 6 different classes of garbage : [ cardboard , glass , metal , paper , plastic , trash]
Once the webcam detects a garbage on the plate , model segregates it using weights (model1.pt) and returns a final value . That value is passed to pi using scp(script.sh) and then servo motor responds(Servo.py) according to the value passed to it .
Also tried segregating by building a tensorflow model , but the model was overfitting the data and as a result validation accuracy was only around 55 % . I will try to improve and update the model.