High-precision three-dimensional (3D) shape measurement technology is essential for improving yield in advanced semiconductor and display device manufacturing processes. Conventional measurement methods such as white-light scanning interferometry and confocal microscopy have been widely used for this purpose. However, these techniques inherently require axial scanning along the optical axis to acquire shape information, and additional spatial scanning is necessary for area measurements rather than point measurements. Consequently, despite their high measurement precision, these methods are difficult to apply to production processes where real-time inspection is required.
In contrast, spectral interferometry enables real-time surface shape measurement without mechanical scanning along the optical axis. Nevertheless, due to its point-based measurement nature, it is limited to acquiring line-profile information along the object transport direction. In this study, we propose a hyperspectral-based real-time three-dimensional shape measurement technique applicable to semiconductor and display manufacturing processes. Because the proposed method is fundamentally based on spectral interferometry, it does not require axial scanning, and it simultaneously acquires line-based data rather than single-point measurements. As a result, the surface shape of objects moving at high speed can be measured in real time.
Furthermore, by implementing a hyperspectral imaging spectrometer with high wavelength resolution, the proposed technique enables precise measurement of surface profiles with large step heights. To verify the reliability of the shape measurement results obtained using the proposed method, certified step-height reference materials provided by the Korea Research Institute of Standards and Science (KRISS) were measured, and agreement within the stated uncertainty relative to the certified values was confirmed. In addition, quantitative performance evaluation was conducted through measurement uncertainty analysis.
The results of this study indicate that the proposed technique can be effectively applied to applications requiring real-time, high-precision shape measurement over large areas, such as wafer-level micro-pattern inspection in semiconductor manufacturing and thickness distribution inspection of insulating layers in secondary battery production.
