Optical character recognition applied to hieratic : sign identification and broad analysis

Abstract

Despite modern advances in digital paleographic techniques, hieratic has largely eluded the thorough application of technological methods for automatic sign recognition. To remedy this, this article presents two new hieratic facsimiles (of the Shipwrecked Sailor and part of the Eloquent Peasant), a novel data set of 13,134 individual hieratic signs, and an Optical Character Recognition (OCR) based program developed to analyze them using an Image Deformation Model. This program is highly accurate and various applications are presented, from single character identification to large-scale sign comparisons. This article provides an important building block in the study of hieratic digital paleography, allowing far more signs to be compared at once than ever before, and offers a free, open-source tool and data set.