An IOT-Based System for Classification and Identification of Plastic Waste Using Near Infrared Spectroscopy

In this paper, an Internet of things (IoT) system for the identification of different types of plastics has been presented using economical near infrared (NIR) spectral sensor. The presented system is highly accurate and economical in distinguishing between various kinds of plastics such as high-density polyethylene (HDPE), polyethylene terephthalate (PET), polystyrene (PS), and polypropylene (PP). Reflectance at four different wavelengths is used to classify and identify these four different plastics. Moreover, multiple machine learning algorithms are applied and compared to improve the accuracy of the system.