Influence of light intensity and magnification on the precision of nanoscale positioning in micro-vision-based dark-field imaging

Zijian Zhu,Xinming Fang,Chenyang Zhao
DOI: https://doi.org/10.1016/j.optlaseng.2023.107753
IF: 5.666
2023-01-01
Optics and Lasers in Engineering
Abstract:Dark-field imaging is widely used in shape analysis and defect detection in micro-vision-based systems because it reflects topography clearly. However, the unclear relationship between dark-field imaging parameters and image quality, along with the lack of dark-field-based image positioning algorithms, has resulted in difficulty combining dark-field imaging with positioning measurements. This paper introduces dark-field imaging into positioning on etched surface patterns. We establish an electromagnetic field model for dark-field imaging and derive an incoherent light point spread function model. Subsequently, the Gaussian fitting grayscale positioning (GFGP) method is proposed for the achievement of pixel-level positioning. Through simulation and physical experiments, we explore the influence of light intensity and magnification on positioning measurement precision and obtain the optimal parameter interval. In the optimal parameter configuration, the GFGP method achieves a precision of 8.2 nm at a speed of 50 Hz. Additionally, the GFGP method assists in restoring image super-resolution, as indicated by experimental results showing a similarity of 0.8714 between the topography under 4 x interpo-lation and the real topography.
What problem does this paper attempt to address?