3D-Printed Object Identification Method using Inner Structure Patterns Configured by Slicer Software
Yuki Kubo, Kana Eguchi, Ryosuke Aoki · 2020
We present an identification method for 3D-printed objects that uses differences in inner structure patterns formed by printing conditions which we can configure in slicer software. Resonant properties change depending on the shape, material, and boundary conditions of the objects, so that our method identifies objects on the basis of differences in resonant properties caused by different inner structure patterns using a machine-learning algorithm. We measured resonant properties as frequency properties using active acoustic sensing. Our method is applicable to 3D-printed objects with a low filling rate while reducing the workload of modeling the inner structure to be used as a tag. To investigate the feasibility of our method, we conducted two experimental evaluations. The results of one showed that our method can identify eight objects with an average classification accuracy of 99.3%.