Due to the increasing demand of machine vision applications in a variety of scenarios, it is necessary to know the capability of the hardware before implementing it. The 1288 Standard by the European Machine Vision Association aims to provide a basis to compare the performance of cameras based on a characterization of the image sensor, using a monochrome light source. This paper aims to investigate the influence the light source has on the measurement results. Which parameters are dependent on it, and which are not? Are there any benefits to using a broadband light source? To answer this question, a series of measurement runs using six different illuminants were performed with the same camera. The illuminants included monochromatic blue, green and red light as well as three different white spectra (CIE E, CIE D65 and white LED). The results show that the influence of the light source on the metrics is limited to the measured quantum efficiency of the camera and related parameters. As a consequence, using a non-monochromatic light source for the measurements might be an option, as it can provide better insight into use-case specific performance and improve comparability.
The accurate measurement of reflectance and transmittance properties of materials is essential in the printing and display industries in order to ensure precise color reproduction. In comparison with reflectance measurement, where the impact of different geometries (0°/45°, d/8°) has been thoroughly investigated, there are few published articles related to transmittance measurement. In this work, we explore different measurement geometries for total transmittance, and show that the transmittance measurements are highly affected by the geometry used, since certain geometries can introduce a measurement bias. We present a flexible custom setup that can simulate these geometries, which we evaluate both qualitatively and quantitatively over a set of samples with varied optical properties. We also compare our measurements against those of widely used commercial solutions, and show that significant differences exist over our test set. However, when the bias is correctly compensated, very low differences are observed. These findings therefore stress the importance of including the measurement geometry when reporting total transmittance.