XBI Dataset

Published: 27 December 2023| Version 1 | DOI: 10.17632/7f7vgps7cn.1
Yangke Li


With the popularization of various electronic devices, the demand for batteries is increasing day by day. However, traditional battery recycling methods need to manually locate batteries from waste electronic devices, which is inefficient and time-consuming. To better alleviate this problem, we propose an efficient solution based on deep learning and X-ray images, which can automatically locate batteries without opening waste electronic devices. Since deep learning models require a large number of training samples, we construct a large-scale high-quality X-ray Battery Inspection (XBI) dataset for battery recycling, which contains 6,000 X-ray images and 23,994 annotated battery instances.



Xi'an Jiaotong University


Object Detection, Battery Recycling