Plastic Segregation Machine Using Audio Signal Analysis

Abstract

In today’s society, plastic detection and segregation is an increasingly important issue due to developing technologies and rapid urbanization. There is a huge problem in plastic segregation which involves manual sorting techniques which are in most cases hazardous, dangerous and effort demanding. As the consumption rate of plastics is multiplying exponentially, it is very important to have an automatic system for plastic segregation. It can be done using various technologies like image processing, IR spectroscopy etc., In this work, improved methods and tools such as machine learning and MFCC (Mel frequency cepstral coefficients) feature extraction are used for the detection of plastic. Based on the audio features, when waste containing metal, wood, paper and plastic is passed through sensing unit, the plastic is detected and automatically sorted by a mechanical setup which will separate the plastic from other waste materials using microcontrollers. The conventional waste management system can be transformed into a smart system by means of automatic plastic segregation system.