Spectral Confocal Microdisplacement Measurement Based on Temporal–Spatial Analysis of Spot Profiles

From chip packaging to microscopic imaging automatic focusing guidance, spectral confocal micromeasurement finds wide applications across various industrial manufacturers. This article focuses on how to precisely extract the central wavelength value of the monochromatic light from the spectral spot images. Thinking from the temporal-spatial conjoint analysis aspect, a novel concept of light intensity gradient consistency (LIGC) is proposed to dynamically evaluate the confidence of any data point in any spot profile from temporal domain with high sampling rate, so then, LIGC-guided temporal features can enable the traditional weighted centroid method survive in the industrial environments suffering with various disturbances. Then, a new method of spectral spot multiprofile weighted centroid (SSMWC) is proposed to acquire more information from spatial domain; by replacing single profile with multiple weighted profiles during the central wavelength calculation procedure from spectral spots, the measurement accuracy is improved notably, as the subjective judgment drawback has been overcome in the traditional single profile calculation method. We name this set of methods (i.e., LIGC and SSMWC) the temporal-spatial analysis of spot profiles (TSASPs). Finally, we develop a microdisplacement measurement prototype with a targeting accuracy of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$1~\mu $ </tex-math></inline-formula>m. The experimental results have verified that the proposed TSASP has promising application prospects in microdisplacement measurement in industrial scenarios, with an average measurement error of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.2470~\mu $ </tex-math></inline-formula>m and with an acceptable response speed of within 14 ms per frame.