Machine-Learning-Inspired Workflow for Camera Calibration.

The performance of modern digital cameras approaches physical limits and enables high-precision measurements in optical metrology and in computer vision. All camera-assisted geometrical measurements are fundamentally limited by the quality of camera calibration. Unfortunately, this procedure is often effectively considered a nuisance: calibration data are collected in a non-systematic way and lack quality specifications; imaging models are selected in an ad hoc fashion without proper justification; and calibration results are evaluated, interpreted, and reported inconsistently. We outline an (arguably more) systematic and metrologically sound approach to calibrating cameras and characterizing the calibration outcomes that is inspired by typical machine learning workflows and practical requirements of camera-based measurements. Combining standard calibration tools and the technique of active targets with phase-shifted cosine patterns, we demonstrate that the imaging geometry of a typical industrial camera can be characterized with sub-mm uncertainty up to distances of a few meters even with simple parametric models, while the quality of data and resulting parameters can be known and controlled at all stages.

A retrospective descriptive analysis of non-physician-performed prehospital endotracheal intubation practices and performance in South Africa.

INTRODUCTION: Prehospital advanced airway management, including endotracheal intubation (ETI), is one of the most commonly performed advanced life support skills. In South Africa, prehospital ETI is performed by non-physician prehospital providers. This practice has recently come under scrutiny due to lower first pass (FPS) and overall success rates, a high incidence of adverse events (AEs), and limited evidence regarding the impact of ETI on mortality. The aim of this study was to describe non-physician ETI in a South African national sample in terms of patient demographics, indications for intubation, means of intubation and success rates. A secondary aim was to determine what factors were predictive of first pass success. METHODS: This study was a retrospective chart review of prehospital ETIs performed by non-physician prehospital providers, between 01 January 2017 and 31 December 2017. Two national private Emergency Medical Services (EMS) and one provincial public EMS were sampled. Data were analysed descriptively and summarised. Logistic regression was performed to evaluate factors that affect the likelihood of FPS. RESULTS: A total of 926 cases were included. The majority of cases were adults (n = 781, 84.3%) and male (n = 553, 57.6%). The most common pathologies requiring emergency treatment were head injury, including traumatic brain injury (n = 328, 35.4%), followed by cardiac arrest (n = 204, 22.0%). The mean time on scene was 46 minutes (SD = 28.3). The most cited indication for intubation was decreased level of consciousness (n = 515, 55.6%), followed by cardiac arrest (n = 242, 26.9%) and ineffective ventilation (n = 96, 10.4%). Rapid sequence intubation (RSI, n = 344, 37.2%) was the most common approach. The FPS rate was 75.3%, with an overall success rate of 95.7%. Intubation failed in 33 (3.6%) patients. The need for ventilation was inversely associated with FPS (OR = 0.42, 95% CI: 0.20-0.88, p = 0.02); while deep sedation (OR = 0.56, 95% CI: 0.36-0.88, p = 0.13) and no drugs (OR = 0.47, 95% CI: 0.25-0.90, p = 0.02) compared to RSI was less likely to result in FPS. Increased scene time (OR = 0.99, 95% CI: 0.985-0.997, p < 0.01) was inversely associated FPS. CONCLUSION: This is one of the first and largest studies evaluating prehospital ETI in Africa. In this sample of ground-based EMS non-physician ETI, we found success rates similar to those reported in the literature. More research is needed to determine AE rates and the impact of ETI on patient outcome. There is an urgent need to standardise prehospital ETI reporting in South Africa to facilitate future research.

Invited review article: measurement uncertainty of linear phase-stepping algorithms.

Phase retrieval techniques are widely used in optics, imaging and electronics. Originating in signal theory, they were introduced to interferometry around 1970. Over the years, many robust phase-stepping techniques have been developed that minimize specific experimental influence quantities such as phase step errors or higher harmonic components of the signal. However, optimizing a technique for a specific influence quantity can compromise its performance with regard to others. We present a consistent quantitative analysis of phase measurement uncertainty for the generalized linear phase stepping algorithm with nominally equal phase stepping angles thereby reviewing and generalizing several results that have been reported in literature. All influence quantities are treated on equal footing, and correlations between them are described in a consistent way. For the special case of classical N-bucket algorithms, we present analytical formulae that describe the combined variance as a function of the phase angle values. For the general Arctan algorithms, we derive expressions for the measurement uncertainty averaged over the full 2π-range of phase angles. We also give an upper bound for the measurement uncertainty which can be expressed as being proportional to an algorithm specific factor. Tabular compilations help the reader to quickly assess the uncertainties that are involved with his or her technique.

Suppression of fundamental-frequency phase errors in phase-shifting interferometry.

I develop and test several new phase-shifting formulas for a Fizeau interferometer, using a spherical cavity of high NA. The associated phase-shift miscalibrations provide a useful range for testing phase-shift error compensation experimentally. After removing the "double-frequency" cyclic phase variation from the error-fringe pattern by well-known methods, the residual phase oscillations occur mainly at 1 and 3 times the fringe frequency. The latter can be addressed by the theory of characteristic polynomials; the former cannot, but is removed by a variation on known averaging methods, and allows empirical optimization of phase-reconstruction performance, which can then be analyzed with characteristic polynomials again.

Null test of an off-axis parabolic mirror. I. Configuration with spherical reference wave and flat return surface.

We demonstrate the precise figure measurement of a one-inch (25.4 mm) diamond-turned 90 masculine off-axis commercial-quality parabolic mirror. The test is carried out with a phase-shifting Fizeau interferometer fitted with a spherical reference surface, auxiliary components and a flat return mirror. We present a detailed and systematic appraisal of the necessary steps for alignment and calibration of the instrument and the alignment of the parabolic mirror. Alignment errors and interferometric sensitivity variations are characterised and corrected, and the results give some insight into the diamond-turning process.

Null test of an off-axis parabolic mirror. II. Configuration with planar reference wave and spherical return surface.

This paper complements our previous study on testing a 25.4 mm diameter diamond-turned 90 masculine off-axis commercial-quality parabolic mirror with a spherical test wave in a phase-shifting Fizeau interferometer (Opt. Express 17, 3196-3210 (2009). In this study I reverse the optical system and use the Fizeau interferometer with a planar reference surface, auxiliary components, and the surface of the transmission sphere as a reflecting spherical return surface. As in the previous paper, I present a description of the necessary steps for alignment and measurement validation. The reversal of the optical system, and associated co-ordinate systems, necessitates some changes of hardware and analysis that provide insight into the underlying symmetries, and may prove useful in a wider context.

Rapid and reliable reference sphere calibration for Fizeau interferometry.

I describe and demonstrate a rapid and simple technique for calibrating spherical reference surfaces in Fizeau interferometry. A flat mirror at the focus of the test wavefront and a partial occlusion of the test beam enable a double-pass measurement of one half of the surface against the other, which gives an error map for half the aperture. Subsequent rotations of the beam stop and stitching together of several measurements yields the full-aperture calibration file. The method always produces point-symmetrical surface maps and is therefore not exact. However, most actual reference surfaces are approximately point symmetric and therefore the trade-off of accuracy for expediency is reasonable in practice.

A deterministic optical figure correction technique that preserves precision-polished surface quality.

A deterministic surface correction technique has been used to improve the surface figure of two fused silica optical flats over a diameter of 60 mm with no measurable degradation in their surface quality at spatial frequencies of

Simultaneous measurement of surface shape and variation in optical thickness of a transparent parallel plate in wavelength-scanning Fizeau interferometer.

Wavelength-scanning interferometry permits the simultaneous measurement of variations in surface shape and optical thickness of a nearly parallel plate. Interference signals from both surfaces of the test plate can be separated in frequency space; however, these frequencies are shifted from the expected frequency by the refractive-index dispersion of the test plate and any nonlinearity that is due to wavelength scanning. Conventional Fourier analysis is sensitive to this detuning of the signal frequency and suffers from multiple-beam interference noise. We propose new wavelength-scanning algorithms that permit a large tolerance for dispersion of the test plate and nonlinearity of scanning. Two 19-sample algorithms that suppress multiple-interference noise up to the second order of the reflectance of the test plate are presented. Experimental results show that the variation in surface shape and optical thickness of a glass parallel plate of 250-mm diameter was measured with a resolution of 1-2 nm rms.

Tunable phase-extraction formulae for simultaneous shape measurement of multiple surfaces with wavelength-shifting interferometry.

The interferometric surface measurement of single or stacked parallel plates presents considerable technical difficulties due to multiple-beam interference. To apply phase-shifting methods, it is necessary to use a pathlength-dependent technique such as wavelength scanning, which separates interference signals from various surfaces in frequency space. The detection window for frequency analysis has to be optimized for maximum tolerance against frequency detuning due to material dispersion and scanning nonlinearities, as well as for suppression of noise from other frequencies. We introduce a new class of phase-shifting algorithms that fulfill these requirements and allow continuous tuning of phase detection to any frequency of interest. We show results for a four-surface stack of nearparallel plates, measured in a Fizeau interferometer.