Year: 2024
Type: article
Abstract: As the current technology of the dental industry changes from analog to digital, this study considered the detection of dental 3D printer output defects as a way to improve the situation in which defects caused by the incorrect output of 3D printers, representative equipment applied with digital dentistry technology, cause waste of materials and time.For the progress of defect detection, You Only Look Once (YOLO) was selected among object recognition algorithms with low memory usage and high learning speed, and a performance comparison test was conducted according to the learning model and setting to find appropriate parameters for the prepared data set.The final learning data established through the test was used to conduct a defect detection test of three types of teeth models (oral, surgical guide, and temporary teeth models).As a result, it was confirmed that defect detection was performed at up to 80%.Since it was confirmed that the detection of defects in the dental model proceeds normally, it was confirmed that this is a significant tool for the detection of defects in dental 3D printers through YOLO.In the future research, research to reduce performance deviations and performance comparison by YOLO version should be conducted, and advanced research that can be tracked in real-time through camera installation in the 3D printer should be conducted using the derived data.
Authors: Dong-Geon Hwang, Doo-Hyun Choi
Institution: Kyungpook National University
Language: English
Subfield: Industrial and Manufacturing Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4402932325