Embedded Files
UV LIGHT
For UV radiation, an 18W fluorescent lamp (Figure a), whose emission spectrum is shown in Figure b, was used.
The first image (up) shows a project drawn by students with markers. By illuminating the painting with the fluorescent lamp, we see that a region stands out (down). By using this method we could see for example whether there is varnish on the painting or it has been removed in some areas.
Visible LIGHT
The artwork used in this case is illustrated in this image. From this particular image, a small area was selected for study, as delimited by the corresponding black box. This area consists of 1402 pixels x 1076 pixels.
The procedure described below, is followed for each filter:
a white surface (photographic paper) is placed in front of the artwork and the corresponding image W is taken and saved,
the white surface is removed,
keeping all settings unchanged, the spectral image A of the project is taken and saved,
total darkness is created and the corresponding B image is taken and saved.
The black image will be used to remove noise from the spectral and white images. The white image will be used to normalize the spectral image. All images are initially converted to black and white (greyscale).
Intensity normalization is performed independently for each filter i but also for each pixel x, according to the equation:
The above calculations are performed using the free software ImageJ (https://imagej.net/ij/index.html), which is a powerful tool for detailed analysis and processing of image files.
Step by step guide for the intensity normalization.
The fraction ranges from zero to one and correlates with the reflectance of the studied area. We notice that as we change the filter, different areas of the image appear brighter. For example, if we have the blue filter (λmean = 531nm) the blue areas of the project appear brighter, while the red areas appear darker. Conversely, for the red filter (λmean = 693nm) the roof of the house appears bright, while the blue areas appear dark.
Finally, we selected four different areas of our project with dimensions of 50 pixels x 50 pixels and we calculated the average of the pixel intensities for each area and for each filter. With these measurements we constructed the graph of the average value of the intensity per wavelength for the four different areas. The graph is shown in the left Figure, while the inset shows the four regions we selected. By comparing the extracted spectrum of each region with other spectrums of standard pigments, the pigments of our artwork could be identified.
INFRARED LIGHT
When the camera is in normal mode, the computer displays the painting as we see it with our eyes (left images). But when the camera is set to night mode, i.e. the infrared radiation is activated, then the draft of the project appears, as shown in right images. Our students were surprised by the result of this process, so the process was repeated several times in different artworks.
Google Sites
Report abuse