Vision-based displacement measurement sensor using modified Taylor approximation approach

The development of image sensors and optics lenses has contributed to the rapidly increasing use of vision-based methods as noncontact measurement methods in many areas. A high-speed camera system is developed to realize the displacement measurement in real time. Conventional visual measurement algorithms are commonly subjected to various shortcomings, such as complex processes, multiparameter adjustments, or integer-pixel accuracy. Inspired from the combination of block-matching algorithm and simplified optical flow, a motion estimation algorithm that uses modified Taylor approximation is proposed and applied to the vision sensor system. Simplifying integer-pixel searching with a rounding-iterative operation enables the modified algorithm to rapidly accomplish one displacement extraction within 1 ms and yield satisfactory subpixel accuracy. The performance of the vision sensor is evaluated through a simulation test and two experiments on a grating ruler motion platform and a steering wheel system of a forklift. Experimental results show that the developed vision sensor can extract accurate displacement signals and accomplish the vibration measurement of engineering structures. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI. [DOI: 10.1117/1.OE.55.11.114103]