However, the calibration operated throughout these applications is often imprecise in relation to the scientific needs for colorimetric quantification [30]. selleckchem A colorimetric calibration is therefore often carried out by combining different polynomial algorithms, multivariate statistics, and neural networks approaches [20]. All of these procedures can successfully reduce the external noise to different extents, pointing out that both the camera settings and its sensor’s response to light play a crucial role for objective color quantification.In this context, the present study introduces a novel colorimetric calibration approach that aims to minimize the effects of the illuminants, camera characteristics and settings. Color image calibration was implemented according to a novel approach: the 3-dimensional sRGB Thin-Plate Spline interpolation (TPS-3D).
The calibration efficiency of this method was compared with the one obtained through the use of a widely used commercial software (i.e., ProfileMaker) as well as with that obtained by multivariate linear regression (Partial Least Squares).2.?Experimental SectionThe Drug_discovery images utilized in this study come from four markedly different operative field and laboratory contexts, in which different devices and lighting conditions occurred or were artificially created.2.1. Calibration/Validation Setup ExperimentIn order to explore the device variability and resolution, two different cameras were used: (i) a commercial high resolution compact Nikon Coolpix P6000, (13.5 real MP-CCD 4.
67�� sensor) with selleck inhibitor optical 4�� NIKKOR lens, providing TIFF 8 bit images (from NRG RAW format) with good macro features (manual white balance control, exposure and metering methods were enabled); (ii) aprofessional high resolution reflex Canon 30D (with a 8.2 real MP CMOS 1.6�� APS-C sensor) with a Canon 10�C22 mm f/3.5�C4.5 ens (used at 22 mm, equivalent to 35.2 mm on a full-frame sensor) providing TIFF 8 bit images (from CR2 RAW format). For both devices, white balance, metering method and exposure were manually defined, while ISO sensitivity was set to the minimum.For each sensor, three consecutive pictures were acquired under four different light conditions: (i) 200 watt Tungsten bulbs (5,000�� K) (T); (ii) 200 watt weakened Tungsten bulbs plus neon tubes plus environmental light (wTNE); (iii) neon tubes plus environmental light (NE); (iv) full sun (i.e., at midday; 6,500�� K) (S). Pictures for the color calibration setup have been taken with three inside altogether different color checkers: the GretagMacbeth ColorChecker SG 140 color-patches, the GretagMacbeth ColorChecker 24 color-patches and the IFRAO Standard ColorChecker 7 color-patches.2.2.