Convolutional Neural Network Based Smart Waste Management System
M Balamurugan, Nandhini S S, E Harine, K M Lathika · 2024
Waste management is becoming more difficult in urban areas due to the substantial amount of trash produced by the area's fast population growth. Traditional systems are expensive and ineffective. To maximize resource usage, a deep learning model and Internet of Things are being integrated into a smart waste management system. introducing bin status monitoring and strengthening waste segregation. The system use EfficientNet which was trained on a dataset comprising paper, cardboard, glass, metal, and plastic, to classify rubbish. This method is used by both pyTorch and the Raspberry pi 5. Trash is recognized by a camera module, sorted by a servo motor into the proper sections on a plastic board, and its fill levels are tracked by an ultrasonic sensor.