Search

In this paper, we focus on the phase noise in optically-pumped magnetometer systems and investigate its sources, and use dual-phase lock-in technique to solve the problem. The results show that the phase noise is maximum when the system operates under zero-magnetic-field condition, and the main source comes from the environment stray magnetic field. The dual-phase lock-in technique improves the phase noise rejection by a factor of 10 compared to the single-phase lock-in technique. The sensitivity of the system is 22 fT Hz−1/2 and the bandwidth is 158 Hz, which offers a favorable outlook for use in clinical settings.

In recent years, phase measuring deflectometry (PMD) has become a popular method to measure the shapes of specular surfaces. However, the inherent slope-height ambiguity problem, which indicates the phase of the recorded fringe pattern is determined by both the height and slope of the test surface, causes considerable ambiguity errors in the measurement results. Existing error reduction methods usually require the initial shape information of the test surface or additional equipment to assist the measurement. These requirements make measurement inflexible in some situations. In this paper, the slope-height ambiguity problem in PMD is modelled and analyzed analytically for the first time, and the system geometry requirements of minimizing the ambiguity error are carried out accordingly. To meet these requirements, a co-axis optical path setup with a telecentric imaging system is proposed. In our simulation and experiment, this system reduces the phase and height errors by up to two orders of magnitude, compared to those from the normally employed system geometry setup.

The ability to accurately estimate the location and geometry of holes is often required in the field of quality control and automated assembly. Projected fringe profilometry is a potentially attractive technique on account of being non-contacting, of lower cost, and orders of magnitude faster than the traditional coordinate measuring machine (CMM). However, we demonstrate in this paper that fringe projection is susceptible to significant (hundreds of µm) measurement artefacts in the neighbourhood of hole edges, which give rise to errors of a similar magnitude in the estimated hole geometry. A mechanism for the phenomenon is identified based on the finite size of the imaging system’s point spread function and the resulting bias produced near to sample discontinuities in geometry and reflectivity. A mathematical model is proposed, from which a post-processing compensation algorithm is developed to suppress such errors around the holes. The algorithm includes a robust and accurate sub-pixel edge detection method based on a Fourier descriptor of the hole contour. The proposed algorithm was found to reduce significantly the measurement artefacts near the hole edges. As a result, the errors in estimated hole radius were reduced by up to one order of magnitude, to a few tens of µm for hole radii in the range 2-15 mm, compared to those from the uncompensated measurements.