AI 虚拟展开千年碳化卷轴首读全文
推荐指数 51.0 NO. 015 · 2026.06.26
发布2026/06/25Score452Comments109
为什么值得看
Vesuvius Challenge 团队用 CT 扫描+机器学习算法，首次完整数字化展开并读取了赫库兰尼姆古卷 PHerc. 1667，无需物理接触。这是两千年来首次
编辑判断
这个项目的真正价值不在考古本身，而是展示了 AI 处理极端噪声、高维结构数据的工程方法论。卷轴碳化后墨水与纸张的对比度极低，团队用 3D 语义分割+纹理分析从体积数据中重建平面文本，这套 pipeline 可以直接迁移到工业无损检测（如电池内部结构分析）和医疗影像领域。
更值得 AI 从业者关注的是其开放协作模式：Vesuvius Challenge 把原始扫描数据、奖金和代码全部公开，用众包竞赛替代传统学术发表周期，两年内把
社区反馈
正面 102 条评论
核心争论：古卷内容是否仅为伊壁鸠鲁哲学，还是包含更广泛的斯多葛学派等失传文献
I am on the vesuvius challenge team that did the segmentation, unwrapping, and ink detection, so feel free to ask any questions.
Given the current rate of progress, how long do you think it will take to decipher the entire collection?
That's a tough one to give a strong estimate of. Some scrolls are easier or harder to unwrap and read for a multitude of different reasons, mostly due to how damaged the scroll was in the eruption, and how easy or not the ink is to read. IIRC from what we've scanned of the herculaneum collection, no