Philippine E-Journals

A Non-Invasive Spectral Imaging System in the Pigment Identification of Watercolor Painting

Jenibel N. Paray | Julian C. Tongson Jr. | Romeric Pobre

 

Abstract:

We demonstrated a non-invasive spectral imaging system in the pigment identification of watercolor paintings. The non-invasive technique used a Digital Light Projector (DLP) and a Sagnac interferometry system that includes an optical module, a cooled CCD camera, and a desktop computer using a Fast Fourier Transform (FFT) computer program to determine the reflectance data of a sampled image. Results show that the non-invasive spectral system can identify pigments with known test swatches with maximum percentage difference of 0.74 %. The reflectance data of the two sample paintings were then statistically computed with known test swatches using t-test to determine the p-value between the known and unknown pigments with test swatches. It is possible to extend this work in which identification of the artist behind the artwork can be done by correlating the p-values of the color pigments with illuminants.