Human versus computer vision in archaeological recording

Abstract

As 3D scanning and photogrammetry are supplanting traditional illustration techniques with increasing speed, archaeologists and architectural historians have sounded alarms about what stands to be lost if hand drawing is altogether eliminated from fieldwork. This paper argues that the most direct threat is to a particular form of archaeological illustration which does not necessarily share the advantages attributed to other kinds of drawing. Recording by means of “technical drawing” communicates a collectively agreed interpretation of the ancient record, and its primary benefit is not stimulating creative thought but rather enhancing human observation. A review of two cases comparing the illustration of ancient Greek architecture through analogue and digital methods indicates that, in practice, both approaches draw attention away from the ancient subject and focus it on distracting protocols for the great majority of the time spent in the field. Even so, technical drawing requires protracted, in-person scrutiny of the subject, whereas 3D technologies pose a genuine risk of altogether eliminating meaningful human interpretation from the recording process. The greater efficiencies of digital techniques suggest a path forward, as time once allocated to tedious stages of technical drawing might be applied toward more thoughtful interpretive tasks. However, such measures must be deliberately integrated into a digital research program through planning around the very different cadences of the digital process.

Mixing Methods

“…Going beyond extensive discussions on the use of 3D data and specifically SfM as a primary documentation method (see e.g., [Anderson 2021;Sapirstein 2021]), this paper presents the digital products and tools that can be created based on SfM and thus aims to demonstrate how these can be an important addition to the standard documentation methods currently in use. These products depict the crucial added value of recording archaeological structures in 3D and can be very diverse (see also Di Giuseppantonio Di Franco et al 2018).…”

Current Digital Archaeology

“…These manifested primarily in a resistance to digital tools that echoed Caraher's (2019) perception of erosion of enskillment and that mourned the loss of by-hand drawing as a way to reflect on and engage with archaeological remains (Morgan et al 2021). Yet Sapirstein (2020) found that the greater efficiencies allowed by the use of digital recording allowed engagement with a more thoughtful recording process (see also Danis 2019).…”

Digitizing an Excavation: A Laser Scanning Database of Maya Architectural Remains

“…Prior to the introduction of digital survey techniques in 2012, architectural documentation in La Blanca was performed through topography and direct measurements recorded in hand drawings. Drawing provides first-hand knowledge of the architectural remains and their interpretation in situ [Morgan and Wright 2018], so manual surveys were not completely discarded in order to ensure an in-person meaningful preliminary study of the remains [Sapirstein 2020] that definitely aids in the analysis of digital data. However, while this traditional survey provides a simplified documentation of the object of study -involves discretization of building geometries and representation of only a number of specific sections, i.e., 2D abstractions of 3D realities [Roosevelt et al 2015]-, digital survey allows its complete documentation through a 3D high-fidelity and accurate model.…”