Global-Scale Archaeological Prospection using CORONA Satellite Imagery: Automated, Crowd-Sourced, and Expert-led Approaches
Global-Scale Archaeological Prospection using CORONA Satellite Imagery: Automated, Crowd-Sourced, and Expert-led Approaches
复制标题
DOI：
10.1080/00934690.2020.1713285
复制
发表时间：
2020-02-20
影响因子：
1.9
通讯作者：
Casana, Jesse
中科院分区：
文献类型：
--
作者：
Casana, Jesse
作为世界上第一个情报卫星计划的一部分，从1960年至1972年收集的解密的CORONA卫星图像提供了近几十年来许多土地使用变化之前的几乎全球性的高分辨率立体图像，因此已被证明是考古研究的非常宝贵的资源。虽然空间校正这些不寻常的全景电影图像所涉及的挑战长期以来一直是研究人员的绊脚石，但一个名为“太阳黑子”的在线工具现在提供了一个简单的过程，可以对CORONA进行有效和准确的正射校正，有助于释放这种历史图像的潜力，用于全球范围的考古勘探。随着这些新的机遇，如何最好地搜索像CORONA提供的大型图像数据集也面临着重大的新挑战。目前考古遥感领域的流行趋势是寻求采用自动化、基于机器学习的方法，或者众包方法来协助识别古代遗址和特征，与此相反，本文主张系统化、集约化和专家主导的“蛮力”方法。一个项目试图绘制北方肥沃新月地区一项大型研究中所有遗址和相关特征的地图，其结果说明了专家主导的分析可能是对复杂考古景观产生细致入微的背景理解的最佳手段。
Declassified CORONA satellite imagery, collected from 1960-1972 as part of the world's first intelligence satellite program, provides nearly global, high-resolution, stereo imagery that predates many of the land-use changes seen in recent decades, and thus has proven to be an immensely valuable resource for archaeological research. While challenges involved in spatially correcting these unusual panoramic film images has long served as a stumbling block to researchers, an online tool called "Sunspot" now offers a straightforward process for efficient and accurate orthorectification of CORONA, helping to unlock the potential of this historical imagery for global-scale archaeological prospection. With these new opportunities come significant new challenges in how best to search through large imagery datasets like that offered by CORONA. In contrast to currently popular trends in archaeological remote sensing that seek to employ either automated, machine learning-based approaches, or alternatively, crowd-sourced approaches to assist in the identification of ancient sites and features, this paper argues for systematic, intensive, and expert-led "brute force" methods. Results from a project that has sought to map all sites and related features across a large study in the northern Fertile Crescent illustrate how an expert-led analysis may be the best means of generating nuanced, contextual understandings of complex archaeological landscapes.