In high-precision optical measuring systems, even the smallest optical elements in the beam path can lead to systematic errors. This also became apparent during the development of the Multi-Aperture Positioning System (MAPS). In measurement series with cover glass on the sensor, systematic measurement deviations were detected depending on the measurement angle, which indicates refraction effects due to the cover glass of the camera.

To eliminate this effect, the cover glass of an industrial camera was removed and comparative measurements were carried out before and after the modification. The measurement results confirmed that the undesirable refraction effects could be eliminated by removing the cover glass, which improved the accuracy of the measuring instrument.

Info about MAPS

The Multi-Aperture Positioning System (MAPS) is an approach for predicting the position of an LED using machine learning in a new type of measurement system. MAPS uses a photogrammetric approach with an aperture mask and a single camera, whereby complex, computationally intensive algorithms were previously used. The aim is to replace these algorithms with a neural network in order to increase the measurement frequency with comparable accuracy.

The neural network was trained with simulated images and tested with real images. It delivered similar results to previously used Gaussian algorithms, but worked faster as no complex image pre-processing is required – which led to a sub-pixel-accurate determination of the light spot center without systematic errors.

A selection of work carried out on this topic can be found here:

Online position correction approach of an industrial robot by using a new photogrammetric measurement system – Luis Garcia, Uwe Bielke, Cornelius Neumann, Theo Hageney, Klaus Banzhaf, Ernst Wiedenmann, Rainer Börret, 2023

Solution to preserve correspondence of measurement and light source using multiple simultaneously with MAPS | IEEE Conference Publication | IEEE Xplore

Machine Learning Based Position Prediction of a Target Tracked by Multi-Aperture Positioning System

Online 6DOF metrology with MAPS utilizing a unique aperture pattern mask

Accuracy evaluation of a new 3D photogrammetric position measurement system for 6D printing

OPUS 4 | Simulation and accuracy evaluation of a new 3D photogrammetric position measurement system

OPUS 4 | Entwicklung einer Simulationsmethodik für das Multi-Aperture-Positioning-System

About the research group

The Multi-Aperture Positioning System (MAPS) was developed as part of a research project at the Center for Optical Technologies (ZOT) at Aalen University. The ZOT is an interdisciplinary research center specializing in innovative applications of optical measurement technology.

The Börret working group involved in the project combines expertise from the fields of geometric optics, image processing and sensor systems in order to realize novel optical instruments for industrial and scientific applications.

Further information on the Center for Optical Technologies (ZOT) and its research projects can be found on the ZOT website.