E-Tadweer Image Dataset (ETID)

Description

The E-Tadweer Image Dataset (ETID) is a curated benchmark dataset developed to train and evaluate computer vision models for automated e-waste classification and recognition. The dataset was constructed by integrating selected publicly available real-world e-waste image datasets with additional images collected using a smart e-waste collection kiosk developed by the authors. Publicly available datasets were systematically reviewed and filtered for suitability for real-world e-waste classification and recognition applications. Selection criteria included the availability of RGB images, public accessibility, relevance to the target e-waste categories, and annotation consistency. Datasets containing non-RGB imaging modalities (e.g., X-ray images), categories outside the intended kiosk application, or incompatible labeling schemes were excluded. The final dataset contains 6,269 RGB images distributed across seven e-waste categories: Battery, Cable, Keyboard, Laptop, Mobile Phone, Mouse, and Printed Circuit Board (PCB). These categories were selected based on their relevance to the operational scope of the developed smart e-waste collection kiosk and represent commonly discarded electronic components encountered in practical recycling scenarios. By combining images acquired from multiple sources with diverse acquisition conditions, including variations in background, viewpoint, object appearance, and imaging environment, the dataset provides a more representative benchmark for evaluating the robustness and generalization capability of computer vision-based e-waste classification models under realistic operating conditions. This dataset is intended to support research on automated e-waste identification, smart recycling systems, and vision-based waste-sorting applications. Researchers can use it to develop, train, and benchmark machine learning and deep learning models for real-world e-waste recognition and classification tasks. The dataset will be updated periodically to include newer types and models of electronic devices, ensuring continued relevance to practical e-waste scenarios.

Institutions

• Qatar University

  Baladīyat ad Dawḩah, Doha

Categories

Funders

• Qatar Research, Development, and Innovation Council

  Grant ID: CCEC01-1029-230097

Licence