Optimizing E-Waste Recycling Through Categorization

The third industrial revolution sparked the development of computers, electronics, and telecommunications. Due to the widespread use of electronic devices and their quick obsolescence, electronic waste, which contains unsafe heavy metals, is becoming an increasingly severe environmental problem. Metal wastes are valuable and has the potential for reuse across various industries. Categorization of such waste will reduce pollution as well as provide economic benefits. That's why we developed an ensemble model based on two CNNs, DenseNet201, and EfficientNetB5. We used our created dataset, which contains images of some common e-waste, like Keyboards, Mobile phones, TVs, Remotes, and Irons. Our proposed model shows an accuracy of 98 % on the dataset, indicating that it can perform the categorization of e-waste with great accuracy.