作者: M.A. Rogerio-Candelera , V. Jurado , L. Laiz , C. Saiz-Jimenez
DOI: 10.1016/J.JAS.2011.04.020
关键词: Computer science 、 Principal component analysis 、 Cultural heritage 、 Mural 、 Artificial intelligence 、 Pattern recognition 、 Painting 、 Digital image analysis 、 Rock art 、 Uncorrelated 、 Archaeology
摘要: Abstract Rock art paintings, and in general mural are one of the many elements cultural heritage complex systems. As different a system have diverse spatial positions, recording allows understanding their interactions. Thus, useful approach to paintings is understand it as microcartography issue, managing each element cartographic coverage. The implemented emphasizes utilization data obtained by remote sensing techniques for extracting kinds information susceptible being analysed, classified plotted differentiate way means possibility reducing redundant Principal Component Analysis (PCA) elaboration false-colour images from uncorrelated bands. A laboratory model was prepared order simulate biodeterioration rock art. samples were photographically recorded thereafter under lighting conditions, PCA applied resulting images. False-colour combining bands allowed us reach results similar those an unsupervised classification. method has been Roman tombs Carmona Necropolis, obtaining good results.