Merged Waste Classification Dataset (MWCD)

Merged Waste Classification Dataset (MWCD) is a benchmark image dataset created by integrating four publicly available waste image datasets: TrashNet, Garbage Classification, GD (Garbage Dataset), and RealWaste. The dataset contains 35,168 RGB images organized into nine standardized waste categories: battery, cardboard, glass, metal, organic, paper, plastic, textile, and trash. Images were merged through label standardization, removal of duplicate and inconsistent categories, and reorganization into a unified directory structure. The dataset is intended for research on waste image classification, transfer learning, computer vision, and explainable artificial intelligence (XAI).