摘要：This article reports on the use of digital image processing techniques to analyze an ultraviolet-fluorescence photograph of the area of the Shroud of Turin where the radiocarbon dating samples were taken. Results of this analysis demonstrate the anomalous nature of the radiocarbon data sample area. The techniques employed in this investigation are those used by the land remote sensing community for the analysis of multispectral imagery. The techniques included extraction of the RGB bands from a digital version of the ultraviolet-fluorescence photograph, use of principal components analysis for statistical compression of the RGB bands, calculation of z-score values using the first component of the principal components analysis, unsupervised classification of the RGB bands via K-MEANS clustering, and image segmentation. These techniques were also used to determine whether the area where the radiocarbon dating samples were taken exhibited any spectral anomalies. A one sample z-test demonstrated that there is a statistically significant difference between the mean z-scores for the four radiocarbon data samples and the mean z-score for a sample area of the Shroud with representative ultraviolet fluorescence. Results of the investigation suggest there is a new set of non-destructive techniques available to support further analysis of the Shroud. Recommendations are made regarding the need for high spatial resolution multispectral imagery and a Shroud coordinate system for mapping Shroud features.
关键词：Brightness values; digital image processing; image segmentation; K-MEANS clustering; principal components analysis; radiocarbon dating sample area; Shroud of Turin; ultraviolet-fluorescence photograph
