Open AccessReal Life Smart Waste Management System [DRY, WET, RECYCLE, ELECTRONIC & MEDICAL]
Author(s) -
A. Radhika
Publication year - 2021
Publication title -
international journal of scientific research in science and technology
Language(s) - English
Resource type - Journals
eISSN - 2395-602X
pISSN - 2395-6011
DOI - 10.32628/cseit2174135
Subject(s) - convolutional neural network , computer science , artificial intelligence , medical waste , object detection , object (grammar) , focus (optics) , deep learning , environmental science , pattern recognition (psychology) , waste management , engineering , physics , optics
The problem of real-life smart waste management system can be solved using automatic waste segregation. In particular, the focus of the article is on the problem of detection (i.e., waste classification). In these 5 classes of waste are taken and segregated them into 5 categories namely dry, wet, recycle, electronic and medical. This system will automatically detect the waste object and segregate it into the respective category. The use of machine learning allowed improving the model with more accuracy. Convolutional Neural Networks (CNN) algorithm which is best used for image classification is used for object detection. The models that was trained are ResNet50, VGG16, InceptionV3 and MobileNetV2. Finally, when compared to the results of all these models, MobileNetV2 has given us the best and highest accuracy of about 98% and 99% respectively.