Simplified white shark with centroid DBN for urban waste management in smart cities

Abstract

Sanitation is becoming more crucial as urban populations rise, making waste management a crucial issue in smart cities. Accurate waste material identification and classification are necessary because the waste items vary as they deteriorate. Furthermore, traditional methods do not overcome the risk of contamination when non-recyclable or hazardous products are improperly disposed of with ordinary waste, and there is an issue of identifying the waste that deteriorates over time. Hence a novel method ''Simplified White Shark with Multivariate Deep Belief Network'' is proposed, it improves system effectiveness and efficiency by identifying waste things that deteriorate with time using an object detection algorithm called Simplified White Shark Object. In addition, a Nearest Centroid Deep Belief Network classification technique for precise waste classification in smart cities is also proposed. The suggested algorithm considerably boosts waste management performance by increasing classification accuracy, efficiency and sensitivity which reaches a maximum of 99.78 %, 99.34 %, 97.69 %, and the minimum error rate of 0.182.

Keywords

• IoT-based smart cities
• Waste management
• Waste material classification
• Waste object detection
• White shark optimizer
• Deep belief network