Page 19 [19]
OCR
darker than the lower section. Figure 1-7(b), an image of a uniformly reflective board taken under the same lighting conditions shows that there is an illumination gradient from dark (top of the image) to light (bottom of the image). A corrected image, where this gradient has been removed (see later for details), produces an image which appears more evenly illuminated, Figure 1-7(c). Optimisation of experimental procedures can minimise the spatial inhomogeneities in illumination which cause these effects. Symmetrical positioning of the radiation sources, avoiding specular reflections and shadows and checking the intensity of the incident radiation around the object with a light meter, are some of the most important factors in achieving these optimal conditions. Further recommendations and solutions for the illumination of objects can be found in the numerous manuals available on the photography of works of art." However, site or object constraints may sometimes impede appropriate positioning of the radiation sources or accurate light intensity measurements, in these cases the inhomogeneities encountered may be addressed via post-processing methodologies. Figure 1-7. (a) Visible-reflected image of a detail from a wall painting fragment from the British Museum (Winged youth from the Tomb of the Nasonii, 1883,0505.5) showing inhomogeneous illumination. (b) Image of a uniformly reflective board taken under the same lighting conditions showing an illumination gradient, from dark (top of the image) to light (bottom of the image). (c) The corrected image appears more evenly illuminated. The mathematical operation of compensating visible, UV- and IR-reflected images for the Spatial inhomogeneities of the radiation source and establishing uniform illumination conditions is known as "flat-fielding" ?? " ? (see Appendix 1 for more explanation). Experimentally this involves recording the distribution of the source on a uniformly reflective board (as seen in the above examples) and dividing the reflected image by the distribution of the radiation in post-processing. This procedure also removes spurious effects caused by variations in the pixel-to-pixel sensitivity of the detector and/or by distortions in the optical path. Version No. 1.0 11 Date : 14/10/2013
