Classification of Garbage For Robtic System Using Deep Learning Techniques

Humans are living in the world where garbage can be seen everywhere. Especially in countries like India garbage is a major problem which leads to Air, Water and Soil pollution. There are many ways to reduce garbage pollution like Disposing waste properly, Recycling, Reusing etc. But you know what garbage can be turned into a useful asset for example Waste to Energy plants which burn solid waste and produce Electricity. This kind of thing is only possible if the waste is separated from one another. Lacking of proper garbage classification system leads to excessive waste of garbage which causes harmful effect to our environment. In that case it is highly required to collaborate with technologies like Convolutional Neural Network (Artificial Intelligent Machine which learns directly from data) which help us to classify waste. InceptionV3 is one of the garbage recognizers which use CNN and Data Augmentation to recognise garbage and sort them. According to sources its accuracy is 99.3 percent. Countries like China, Columbia and Indonesia is also working toward garbage recognizer machine which is fruitful for the environment as well as for the peoples living here. There are many techniques through which can be used in the robotic systems to segregate the waste accordingly using the machine learning and deep learning techniques. The main aim of the paper is to compare the deep learning techniques and prove which is the best technique that can be used for a particular system.