Visual Servoing and Deep Capsule Network Learning for Contactless Smart Waste Segregation

Due to the rapid urbanization and industrialization, lots of waste is generated but not properly disposed. Domiciliary wastes are also dumped either in the public places or waste land leading to various pollutions. This leads to the growth of micro-organisms which cause harmful diseases. These wastes are picked up and segregated manually by sanitation workers which may later affect their health. This paper presents a smart waste segregation model that can distinguish and separate the waste into degradable, non-degradable and also high radioactive and low radioactive waste materials. The automated waste segregator uses sensors as well as visual servoing and deep capsule network learning technologies so that human intervention is completely avoided. It ensures accurate segregation of waste into categories like paper, plastic, metal, non-metal and uncategorizable wastes. These waste products are further sent for recycling or incinerated without eliminating any hazardous gases into the atmosphere. This model is extremely beneficial during pandemic situations for contactless waste segregation as well as in safe handling of hospital waste.