Displacement measurement of the compliant positioning stage based on a computer micro-vision method

Abstract

We propose a practical computer micro-vision-based method for displacement measurements of the compliant positioning stage. The algorithm of the proposed method is based on a template matching approach composed of an integer-pixel search and a sub-pixel search. By combining with an optical microscopy, a high resolution CCD camera and the proposed algorithm, an extremely high measuring precision is achieved. Various simulations and experiments are conducted. The simulation results demonstrate that the matching precision can reach to 0.01 pixel when the noise interference is low. A laser interferometer measurement system (LIMS) is established for comparison. The experimental results indicate that the proposed method possesses the same performance as the LIMS but exhibits a greater flexibility and operability. The measuring precision can theoretically attain to 2.83 nm/pixel.

A Review of Computer Microvision-Based Precision Motion Measurement: Principles, Characteristics, and Applications

“…By directly operating on the pixel values, it does not require any form of specific target pattern, and it is relatively robust to image noise since highly redundant information is involved. Hence, template matching approaches can work well even with few image features, thereby are popular and have been widely adopted for motion measurement in microvision systems [77]- [80]. However, the high computational cost might limit the applications for real-time motion tracking.…”

“…Besides, the range-to-resolution ratio can evaluate the multiscale measurement ability of a method. Therefore, in Figure 16, we review and summarize different state-of-the-art motion measurement methods [31], [45], [77], [82], [105], [109], [113], [122], [123], [149]- [165] by the range-to-resolution ratio in terms of Eq. 18 and different motion measurement algorithms, which can more comprehensively investigate the motion measurement performances.…”

“…In works [45], [77], [82], [113], [122], [155], [160], [161], [165], subpixel algorithms have been applied, which significantly improve the measurement accuracy and contribute to a high range-to-resolution ratio with a trade-off of measurement efficiency. Since the smallest unit of an image is one pixel, displacements are normally measured at integer-pixel resolution.…”

“…The geometric approach [155], including the centroid method and gray-gravity method, utilizes the geometric characteristics of a target in the image to obtain subpixel measurement data. The moment approach constructs the image intensity into the moment feature descriptors [77], and it has a strong anti-noise ability, as described in Section II-C2. Interpolation approach is also a popular choice.…”

“…Only a limited number of works for 1-DOF visual measurement are proposed, which mainly exploring high-frequency rate in one dimension [170]. 2-DOF measurement, on the other hand, has been widely studied [72], [77], [113], [171], where in-plane position information is sufficient for the task requirement. All the tracking techniques mentioned in Section II-C are able to easily perform motion measurement for in-plane displacement by locating the target in the image sequence.…”