Flexible Calibration Method and Application of Passenger Car HUD Detection Based on Collaborative Robot
Ding Zhigang,Yanlu Lv,Linghua Kong,Zhiming Dong,Jishi Zheng,Jiadi Zhang,Jiaxin Liu
DOI: https://doi.org/10.1109/access.2024.3350631
IF: 3.9
2024-03-12
IEEE Access
Abstract:At present, the HUD calibration of the whole vehicle enterprise adopts the traditional physical Master Gauge to calibrate the camera, and the scheme has low flexibility, occupies a large space, and is unsuitable for the mixed production of different types of vehicles. A flexible calibration approach is presented based on collaborative robots for HUD inspection of passenger cars. They are considering that the general HUD (Head-Up Display) virtual image distance is 2–3 m. A 2.3 m hand-eye calibration method is investigated by mounting a monocular camera on a collaborative robot, considering the effect of ambient light on long-distance hand-eye calibration, and employing customized alumina calibration plates and lighting through appropriate angles. Using the area_scan_polynomial model for camera calibration, the camera's internal reference is solved, and the average error of the resulting camera's internal reference is about 0.15. Compared with the master gauge camera calibration, it shows higher flexibility. The hand-eye calibration results show that the Translation part's maximum error is about 1mm, the maximum error of the Rotational position is about 0.1 degree, and the camera calibration error is within 0.08 pixels. For repeatability, the median of Translation RMS was 0.5 with an interquartile range of 0.171, and the median of Translation Maximum was 0.972 with an interquartile range of 0.216. The median of Rotation RMS was 0.0325 with an interquartile range of 0.0125. The median of the Rotation Maximum was 0.059, with an interquartile range of 0.016. The data of translation is relatively scattered, with an extensive range of changes, but it still maintains a certain repeatability. The Rotation data is relatively stable, with a small degree of change and good repeatability, further meeting the use requirements in the final inspection HUD scene of the whole vehicle.
computer science, information systems,telecommunications,engineering, electrical & electronic