Waste Classification Based on Computer Vision Using Deep Learning Models and Public Awareness

Addressing the escalating challenge of global waste management, this research introduces an innovative mobile application utilizing deep learning models for efficient waste classification. The conventional manual sorting processes prove laborious and time-intensive, leading to inefficiencies in waste management systems. This solution employs computer vision and machine learning algorithms to identify and categorize waste based on visual cues such as color, texture, and shape. The system includes an Android app to promote public awareness and a web dashboard for intelligent waste management. Evaluation of three deep learning models-MobileNetV3Large, InceptionV3, and ResNet50-using a substantial waste image dataset reveals initial test set accuracies of 78%, 82%, and 76%, respectively. Following web scraping and misclassification elimination, significant accuracy improvements were observed: MobileNetV3 increased to 82%, InceptionV3 to 89%, and ResNet50 to 90%. These findings underscore the EcoQuest Application’s efficacy in enhancing waste management practices, contributing to a more sustainable and environmentally conscious future.