Computer vision methods and rock art: towards a digital detection of pigments
- PDF / 1,419,761 Bytes
- 13 Pages / 595.224 x 790.955 pts Page_size
- 23 Downloads / 168 Views
ORIGINAL PAPER
Computer vision methods and rock art: towards a digital detection of pigments Enrique Cerrillo-Cuenca · Pedro Ortiz-Coder · ´ Jos´e-Angel Mart´ınez-del-Pozo
Received: 19 February 2013 / Revised: 29 May 2013 / Accepted: 3 June 2013 © Springer-Verlag Berlin Heidelberg 2013
Abstract A non-invasive procedure for assessing and interpreting the pigments of rock art paintings through computer vision and photogrammetric techniques is presented. The method is designed to document and interpret poorly preserved pigments by making use of advanced techniques of photogrammetry and computational imaging. Two different software solutions that were developed by the authors have been used for this purpose. Finally, two case studies of rock art paintings from Spain have been analysed, which show the reliability of the method. Keywords Rock art paintings · Computer vision · Pigments · Statistical imaging · Photogrammetry
Introduction The digital documentation of rock art paintings is a continuously evolving aspect of fieldwork in prehistoric archaeology. The need to study paintings using non-direct contact methods has caused computational imaging to be one of the preferred documentation techniques for archaeologists. Highly detailed 3D models are necessary for generating complete documentation of any archaeological feature (Plets et al. 2012). Accurate geometric and radiometric information permits us to preserve the attributes of surfaces through time and to use different digital tools to study it in an optimal way before a deterioration issue can occur. This concern can be found in many of the papers that have E. Cerrillo-Cuenca () · P. Ortiz-Coder · ´ Mart´ınez-del-Pozo J. A. Archaeology Institute of M´erida, Plaza de Espa˜na, 15, 06800 M´erida Badajoz, Spain e-mail: [email protected]
addressed the application of digital techniques to rock art recording (Plets et al 2012; Cai 2011; Portillo et al. 2008; D´ıaz-Andreu et al. 2006). In spite of the absence of direct manipulation, it is clear that digital recording of rock art paintings can enhance the recognition of less visible features or motifs, which can often be misleading when using direct tracings (Plets et al 2012; Cai 2011; Mark 2002; Mark and Billo 2006) in areas that are more frequently exposed over eroded surfaces (Brady and Gunn 2012). At the same time, computer vision techniques have evolved toward a more efficient and accessible set of tools that can improve the recording and recognition of archaeological features, for which rock art is one of the key fields (Mudge et al. 2012). From our point of view, in the case of rock art paintings, a robust digital documentation method should speed up the recording of panels; however, this approach is also a key opportunity to (1) provide accurate metric information about the elements that are represented in the image, (2) boost the recognition of poorly preserved paintings and (3) decrease the degree of subjectivity when interpreting the paintings. Computer vision can offer a significant number of libra
Data Loading...
