A Tampering Features Dataset of Fuel Dispenser

This dataset provides high-resolution images and precise object detection annotations of assembled printed circuit boards (PCBs) from fuel dispensers, designed to support computer-vision-based research on illicit metering detection and anti-fraud inspection.The dataset contains 189 high-resolution images of fuel dispenser PCBs seized during real-world law enforcement operations, covering 5 mainstream fuel dispenser brands.

To eliminate perspective distortion, all images underwent rigorous lens distortion correction and four-point homography transformation preprocessing. Six categories of tampering features — including the addition of tampered surface-mount resistors and other concealed circuit modifications — are manually annotated in YOLO format and cross-validated by domain experts.