Battery Image Dataset for EV Circularity Research (19 Battery Types)

This dataset contains annotated images and videos of electric and hybrid vehicle batteries from 19 distinct vehicle types. The data is intended for computer vision research, including image classification, object detection, and representation learning related to battery identification and circular economy applications.

The dataset was created within the RECIRCULATE project, a European research initiative focused on circular economy approaches for electric vehicle batteries.

Images originate from multiple sources:

- "Tesla Model 3" and "Tesla Model S black" images are derived from the MTech project EV battery pack dataset (CC BY 4.0).

- Some Ford Transit images are original photographs captured by the dataset author at Centria’s laboratory. Other Ford Transit images and videos are captured by other contributors within the RECIRCULATE project.

- All remaining images and videos were captured by contributors within the RECIRCULATE project.

Version 2 expands the dataset from 3 to 19 battery classes, updates the dataset title accordingly, and distributes all images at their original resolution. Data augmentation has been applied to selected classes to address class imbalance.

See the README file in the archive for full class listings, directory structure, augmentation details, and attribution information.