Integrating InceptionV4 and Color Histogram for Scalable E-Waste Classification Framework

The increasing volume of electronic waste poses severe environmental challenges, necessitating efficient and automated classification methods. This study proposes a hybrid approach combining InceptionV4 for deep hierarchical feature extraction and color histograms for encoding color distributions. The fused features are optimized using Principal Component Analysis and classified using a Random Forest Classifier. Comprehensive evaluations on the E-Waste Image Dataset demonstrate the method's effectiveness, achieving high accuracy, precision, and F1 scores while addressing the limitations of single-feature approaches. This scalable framework offers significant advancements in automated e-waste classification, contributing to sustainable waste management and environmental conservation.