DIGITAL IMAGE ANALYSIS OF THE VISIBLE REGION THROUGH SIMULATION OF ROCK ART PAINTINGS
Non-destructive rock art recording techniques are getting special attention in the last years, opening new research lines in order to improve the level of documentation and understanding of our rich legacy. This paper applies the principal component analysis (PCA) technique in images that include wavelengths between 400-700 nm (visible range). Our approach is focused on determining the difference provided by the image processing of the visible region through four spectral images versus an image that encompasses the entire visible spectrum. The images were taken by means of optical filters that take specific wavelengths and exclude parts of the spectrum. Simulation of rock art is prepared in laboratory. For this purpose, three different pigments were made simulating the material composition of rock art paintings. The advantages of studying the visible spectrum in separate images are analysed. In addition, PCA is applied to each of the images to reduce redundant data. Finally, PCA is applied to the image that contains the entire visible spectrum and is compared with previous results. Through the results of the four visible spectral images one can begin to draw conclusions about constituent painting materials without using decorrelation techniques.