Waste segregation using deep learning algorithm

Abstract

In 2017, India is in 177th position of the Green ranking in World Economic Forum. Due to poor handling of air pollution and waste management, India has moved from 141st position to 177th position. With the emerging smart city development across the cities in India, Smart Garbage Management system is the need of the hour. It is estimated that the generated waste is more than 2.0 billion tones. The existing way of garbage management system in India involves waste collection from homes and industries and dumping into dump yards. The segregation of solid waste is completely done by manual laborers which is less efficient, time-consuming and not completely feasible due to large amount of waste. This paper proposes an automated waste classification system using Convolution Neural Network (CNN) algorithm, a Deep Learning based image classification model used to classify objects into bio and non-biodegradable, based on the object recognition accuracy in real-time. This algorithm is suitable for a large amount of waste segregation process. Python index package of spyder is used to identify and classify the waste material in real-time through webcam. In this paper, the first phase of the waste segregation process is carried out where initially the system is able to detect the object provides the relative match percentage of each object. Open source software libraries such as Tensor flow and Spyder is used for this process.