Learned Anomaly Detection with Terahertz Radiation in Inline Process Monitoring

Abstract

Terahertz tomographic imaging as well as machine learning tasks represent two emerging fields in the area of nondestructive testing. Detecting outliers in measurements that are caused by defects is the main challenge in inline process monitoring. An efficient inline control enables to intervene directly during the manufacturing process and, consequently, to reduce product discard. We focus on plastics and ceramics, for which terahertz radiation is perfectly suited because of its characteristics, and propose a density based technique to automatically detect anomalies in the measured radiation data. The algorithm relies on a classification method based on machine learning. For a verification, supervised data are generated by a measuring system that approximates an inline process. The experimental results show that the use of terahertz radiation, combined with the classification algorithm, has great potential for a real inline manufacturing process. In a further investigation additional data are simulated to enlarge the data set, especially the variety of defects. We model the propagation of terahertz radiation by means of the Eikonal equation.