Model Registration Techniques for 3D Optical Scanning
Dr. Trevor Lujan
3D optical scanning can be used to visualize morphological differences in objects, including mechanical wear in biomaterials. To effectively evaluate these deviations, two 3D models must be analyzed. This is done by registering the models together using unique geometric features shared between the models then calculating surface deviation through colorimetric mapping. One registration process is to use unique geometric markers shared between the two models. However, no standard established geometric marker currently exists for this registration which leads to variation between research groups using different markers and software. Therefore, the objective of this study was to determine a standardized geometric marker for registration purposes. Based on our previous work, a tibia bone geometry was chosen to act as our model. The geometric markers were simple cubes of varying sizes (0.13, 1.0, 8.0 cm3). A 3D optical scanner was used to generate 3D models of the tibia with the geometric marker, following our previously established methods. This was done in triplicate for each cube size. The surface variation between the two models was then visualized through 3D colorimetric maps generated within the scanning software. Surface variation was analyzed for the cube and the top surface of the tibia within the central, medial, and lateral regions. Surface variation in the cube linearly decreased as block size increased (R2 = 0.99), with the 8 cm3 cube having the lowest surface variation (4.7 ± 2.0 µm). The 8 cm3 cube also had the lowest average surface variations across the tibia surface proving that an 8 cm3 cube is the best choice for registration. This is the first study to provide a repeatable geometric marker for accurate registration of two 3D models. By standardizing the registration process, this study can advance the use of 3D optical scanning for evaluation between 3D models.
Burruel, Dylan; Lujan, Trevor; and Benfield, Kate, "Model Registration Techniques for 3D Optical Scanning" (2022). 2022 Undergraduate Research Showcase. 79.