PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Analyzing Ancient Maya Glyph Collections with Contextual Shape Descriptors
Edgar Roman-Rangel, Carlos Pallan, Jean-Marc Odobez and Daniel Gatica-Perez
International Journal of Computer Vision 2010.


This paper presents an original approach for shape-based analysis of ancient Maya hieroglyphs based on an interdisciplinary collaboration between computer vi- sion and archeology. Our work is guided by realistic needs of archaeologists and scholars who critically need support for search and retrieval tasks in large Maya imagery col- lections. Our paper has three main contributions. First, we introduce an overview of our interdisciplinary approach to- wards the improvement of the documentation, analysis, and preservation of Maya pictographic data. Second, we present an objective evaluation of the performance of two state-of- the-art shape-based contextual descriptors (Shape Context and Generalized Shape Context) in retrieval tasks, using two datasets of syllabic Maya glyphs. Based on the iden- tification of their limitations, we propose a new shape descriptor named Histogram of Orientation Shape Context (HOOSC), which is more robust and suitable for descrip- tion of Maya hieroglyphs. Third, we present what to our knowledge constitutes the first automatic analysis of visual variability of syllabic glyphs along historical periods and across geographic regions of the ancient Maya world via the HOOSC descriptor. Overall, our approach is promising, as it improves performance on the retrieval task, has been suc- cessfully validated under an epigraphic viewpoint, and has the potential of offering both novel insights in archeology and practical solutions for real daily scholar needs.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Machine Vision
ID Code:7135
Deposited By:Jean-Marc Odobez
Deposited On:05 March 2011