A Hand-Eye Calibration Method for Skin in Situ Printing Based on Feature Point Extraction by Point Cloud Segmentation

Skin burn treatment is an urgent problem in the clinic. Skin in situ printing technology was expected to treat skin burns with accurate in situ construction of skin tissue. The first step in the skin in situ printing is to calibrate the hand-eye of the printhead and scanner. However, the accuracy of the existing industrial hand-eye calibration methods is difficult to be applied to the treatment of burned skin. According to the high precision requirement of skin in situ printing, a hand-eye calibration method based on point cloud segmentation feature point extraction is proposed in this paper. Firstly, the statistical filtering method, cluster segmentation method, and RANSAC method are used to remove noise points, plane fitting, and feature point extraction for the scanned point clouds. Then, the feature points processed by the point cloud were matched with the feature points in the coordinate system of the three-dimensional measurement platform and the rotation translation matrix between the coordinate system of the three-dimensional printing platform. The coordinate system of the binocular laser three-dimensional scanner was obtained by the singular value decomposition method. Finally, the wrong cloud data is removed in turn. The hand-eye calibration experimental results show that compared with the method without feature point extraction by point cloud segmentation, the proposed method reduces the average distance and maximum error of the XYZ axis by 78.477% and 82.301%, respectively. Compared with other hand-eye calibration methods, the average distance error is increased from 0.060mm to 0.040mm. It is proved that the hand-eye calibration based on feature point extraction by point cloud segmentation can meet the needs of skin in situ printing, which makes preparation for subsequent skin in situ printing experiments and makes it possible to treat skin burns in situ clinically.