The 2024 Autumn Conference of the Korean Society for Precision Engineering was held from Wednesday, November 13 to Friday, November 15 in Gyeongju, a UNESCO City of Historical and Cultural Heritage. The conference featured oral and poster presentations across 11 divisions, including precision machining, precision measurement, and robotic control & automation. Meterlab presented on its various measurement technologies in the precision measurement session.
Figure 1. Oral presentation: “High-speed thickness and profile measurement using a hyperspectral camera” (Senior Researcher Heulbi An)
First, Senior Researcher Heulbi An delivered a presentation titled “High-speed thickness and profile measurement using a hyperspectral camera.” The talk began with an overview of Meterlab and introduced the spectrometers and sensors developed in-house for high-speed thickness and shape measurements. In particular, it covered the s-Nova-H hyperspectral camera series—developed by Meterlab with a focus on thickness measurement—and the measurement technologies and sensor products based on it. The presentation then introduced an approach that combines a hyperspectral camera with a spectral interferometer to measure surface profiles, and showed results including thickness profiles of glass/film areas and surface profile measurements of step-certified standard specimens.
Figure 2. Oral presentation: “Improving the performance of deep-learning algorithms for thin-film thickness analysis by accounting for reflectance fluctuation” (Researcher Junyoung Lee)
Next, Researcher Junyoung Lee presented on “Improving the performance of deep-learning algorithms for thin-film thickness analysis by accounting for reflectance fluctuation.” The talk introduced cases applying high-performing deep-learning algorithms to measurement tasks. In particular, it explained limitations encountered during dataset construction for training and proposed a method of incorporating datasets that reflect real measurement conditions to address these limitations. Results demonstrating improved algorithm performance were presented.
Figure 3. Oral presentation: “Uncertainty evaluation for multilayer thin-film thickness measurement using spectroscopic reflectometry” (Senior Researcher Jaeseok Bae)
Finally, Senior Researcher Jaeseok Bae presented on “Uncertainty evaluation for multilayer thin-film thickness measurement using spectroscopic reflectometry.” While reliable measurements require reporting measurement uncertainty alongside analysis results, there has been a lack of established methods for uncertainty evaluation in multilayer thin-film measurement. To address this, he proposed a new method that evaluates multilayer film thickness measurement uncertainty by tracing measurements back to a length standard using single-layer certified reference materials and single-layer specimens produced concurrently with the multilayer samples.
Through this conference, Meterlab presented a range of research outcomes—from measurement technologies using hyperspectral cameras to applications of deep learning algorithms and methods for measurement uncertainty evaluation. Meterlab will continue to pursue technological innovation and R&D to provide more accurate and reliable measurement solutions. We are committed to contributing to the advancement of precision measurement technology in Korea and delivering greater value to our customers. Thank you.
