Micro-vision Super-Resolution Restoration and Positioning Based on Ultra-Precision Machining Topography Guidance

Zijian Zhu,Chenyang Zhao,Yueping Xi
DOI: https://doi.org/10.1109/tim.2024.3470255
2024-01-01
Abstract:Micro-vision-based precision positioning measurement technology faces limitations due to optical diffraction, making it challenging to acquire high-resolution images. This limitation fundamentally affects the positioning resolution and accuracy. Traditional super-resolution restoration methods only consider the grayscale distribution of the image, failing to restore the detailed features blurred by lens diffraction. To address this issue, this article proposes a micro-vision super-resolution restoration based on ultraprecision machining topography guidance (SRT), aiming to achieve high-resolution imaging and precision positioning. The method involves developing a microstructure surface with known topography using ultraprecision machining technology, exploring the imaging mechanism of microstructure surface topography, and realizing super-resolution restoration and interpolation of microscopic images based on topography features. In the experiment, anchor microstructures and chisel microstructures are processed through ultraprecision machining for micro-imaging and positioning measurement. The experimental results demonstrate that the SRT method effectively restores the image feature information and achieves a positioning accuracy of 20.7 nm using a simple traditional image registration algorithm. It improves the microscopic precision positioning measurement from the image level and has potential application value in industrial inspection and nanoscale positioning measurement.
What problem does this paper attempt to address?