DIN/EN ISO 10360-10: Performance Testing for Optical 3D Measuring Systems

Probing error (P_Form) is the range of residual deviations from a least-squares sphere fit to all acquired data points on a calibrated reference sphere. The sensor acquires a defined minimum number of points across the sphere surface; a least-squares algorithm fits an ideal sphere to this point cloud; P_Form is the difference between the maximum and minimum residual deviation.

Formally:

PForm=dmax−dminPForm​=dmax​−dmin​

where dmaxdmax​ is the largest positive residual and dmindmin​ is the largest negative residual from the best-fit sphere surface.

P_Form quantifies the system’s ability to reproduce the form of a known geometric primitive.A low P_Form value indicates that the sensor captures surface geometry with high local consistency and low noise.

Sphere-spacing error (E_uni) is the absolute difference between the measured centre-to-centre distance of 2 calibrated reference spheres and the traceable reference value of that distance.

Formally:

Euni=∣Lmeasured−Lreference∣Euni​=∣Lmeasured​−Lreference​∣

where LmeasuredLmeasured​ is the centre-to-centre distance determined from the point cloud and LreferenceLreference​ is the calibrated value from the artefact certificate.

E_uni quantifies the system’s length measurement accuracy across the measurement volume. E_uni is the volumetric equivalent of the length measurement error used in tactile CMM testing under ISO 10360-2. A single E_uni test result covers one spatial orientation and one separation distance; the complete test protocol requires measurements at multiple positions and orientations within the measurement volume.

The measurement volume is the 3-dimensional working space within which a measuring system produces results that comply with its manufacturer-stated specifications. ISO 10360-10 requires that performance tests sample the measurement volume systematically — not only at the centre or at a single working distance.

Systematic volume sampling detects 3 categories of spatially dependent errors: edge-region distortion from lens or laser geometry, depth-dependent scale errors from calibration nonlinearity, and position-dependent noise from varying illumination intensity across the field.

ISO 10360-10 distinguishes between 2 acquisition modes. Single-scan acquisition captures a test artefact in one exposure or one scan pass without repositioning the sensor or the artefact. Multi-scan acquisition combines multiple exposures or scan passes — with or without sensor repositioning — to produce the final measurement result.

Both modes are valid under ISO 10360-10, provided the manufacturer declares which mode applies to the stated MPE values. Multi-scan results are not directly comparable to single-scan resultsfor the same performance metric.

First, the sphere diameter must be large enough relative to the sensor’s spatial resolution to acquire the minimum required number of data points across the sphere surface. Second, the sphere surface finish must produce a stable, diffuse reflection consistent with the sensor’s illumination wavelength — surfaces with high specularity or wavelength-dependent absorption produce systematic errors in the point cloud. Third, the sphere form deviation (calibrated) must be smaller than the probing error to be measured, so the artefact does not dominate the test result.

ISO 10360-10 requires sphere placement at multiple positions and orientations within the measurement volume. A minimum of 5 sphere positions — distributed to sample the near, far, central, left, and right regions of the field — detect spatially dependent performance variations. Each position yields one P_Form result; paired sphere positions yield E_uni results for each separation distance and orientation tested.

Optical sensors respond to the optical properties of the measured surface — reflectance, specularity, colour, and translucency — in ways that tactile sensors do not. A reference sphere with a matte white ceramic coating produces a fundamentally different signal than the same sphere with a polished steel surface, even when their geometry is identical.

ISO 10360-10 accounts for this dependency by requiring that test artefacts match — or are representative of — the reflectance class of the workpieces the system is intended to measure. This artefact-surface matching requirement has no equivalent in tactile CMM testing under ISO 10360-2.

P_Form is calculated in 4 steps.

A P_Form result exceeds the manufacturer’s MPE when the range of residuals is larger than the declared probing error limit. Exceedance at any single test position constitutes a non-conforming result for that test.

E_uni is calculated as Euni=∣Lmeasured−Lreference∣Euni​=∣Lmeasured​−Lreference​∣. The centre of each sphere is determined by least-squares sphere fitting to the acquired point cloud for that sphere.

E_uni is evaluated independently for each pair of sphere positions and each spatial orientation tested. The complete E_uni test result is the maximum E_uni value recorded across all tested positions and orientations within the measurement volume.

MPE (Maximum Permissible Error) is the manufacturer-declared upper limit for a performance metric under specified test conditions. ISO 10360-10 does not define universal MPE values; each manufacturer specifies the MPE applicable to each product and each operating mode.

Acceptance testing determines whether a specific measuring system produces P_Form and E_uni results that do not exceed the MPE values declared for that system. A single exceedance of any MPE value constitutes a non-conforming acceptance test result, regardless of results at other positions.

Measurement uncertainty, evaluated in accordance with the Guide to the Expression of Uncertainty in Measurement (GUM / ISO/IEC Guide 98-3), applies to both the reference artefact calibration and the test measurement results. Uncertainty is treated separately from MPE and does not substitute for it.