Application of Computational Imaging and Artificial Intelligence in the Recovery of Carbonized Herculaneum Papyri

計算成像與人工智慧在恢復碳化赫庫蘭尼姆莎草紙中的應用


Introduction

Researchers have successfully utilized high-resolution 3D scanning and artificial intelligence to read carbonized scrolls from the library of Herculaneum without physical intervention.

研究人員成功利用高解析度 3D 掃描與人工智慧,在無需物理干預的情況下,讀取來自赫庫蘭尼姆圖書館的碳化卷軸。

Main Body

The Herculaneum scrolls, carbonized during the 79 AD eruption of Mount Vesuvius, have remained largely illegible since their 1752 discovery due to the fragility of the charred papyri and the visual similarity between the ink and the substrate. Historically, physical attempts to unroll these documents frequently resulted in structural degradation. The Vesuvius Challenge has mitigated these risks through the implementation of particle accelerator scans and the development of AI algorithms designed for 'virtual unwrapping.' These computational models are currently calibrated to individual scrolls to account for variations in ink composition, though the objective is to achieve a generalized model capable of universal ink detection across the collection.

赫庫蘭尼姆卷軸在西元 79 年維蘇威火山爆發期間被碳化,由於燒焦的莎草紙極其脆弱,且墨水與底材在視覺上非常相似,自 1752 年發現以來大多無法閱讀。在歷史上,嘗試物理性地展開這些文件經常導致結構損壞。「維蘇威挑戰」透過實施粒子加速器掃描並開發專為「虛擬展開」設計的 AI 演算法,降低了這些風險。目前的計算模型針對個別卷軸進行校準,以適應墨水成分的差異,但目標是實現一個能夠在整個收藏中通用偵測墨水的通用模型。

Recent achievements include the extraction of 1.5 meters of text from a scroll core dating to the 2nd century BC. The content pertains to ethics, art, and human nature, with references to the nephew of Chrysippus, leading researchers to identify the Stoic philosopher as the probable author. Additionally, the recovery of 70 columns from 'On Vices, Book 1' and the expansion of Philodemus's 'On Gods' to at least an eight-book series have been documented. These findings are significant as they provide primary source texts, thereby reducing reliance on the potentially biased summaries provided by subsequent ancient authors. The project continues to scale, with the recent transition from partial column recovery to the full reading of entire scrolls.

最近的成就包括從一卷可追溯至公元前 2 世紀的卷軸核心中提取了 1.5 公尺的文本。內容涉及倫理、藝術與人性,並提及克律西波斯的侄子,使研究人員認定這位斯多葛派哲學家可能是作者。此外,紀錄顯示已恢復《論惡行》第一卷的 70 欄文本,且費洛德姆斯的《論神》已擴展至至少八卷系列。這些發現具有重要意義,因為它們提供了第一手原始文本,從而減少了對後世古代作者可能帶有偏見的摘要之依賴。該項目持續擴大規模,近期已從恢復部分欄位過渡到完整閱讀整卷卷軸。

Conclusion

The integration of AI and advanced imaging has transitioned the project from technical validation to the systematic recovery of lost philosophical texts.

AI 與進階成像技術的結合,使該項目從技術驗證階段轉型為系統性地恢復失傳的哲學文本。

Vocabulary Learning

The Architecture of 'Nominalization' and the C2 Passive

To bridge the gap from B2 to C2, a student must move beyond describing actions and begin constructing concepts. The provided text is a masterclass in conceptual density, achieved primarily through high-level nominalization—the process of turning verbs and adjectives into nouns to create a formal, objective academic tone.

◈ The Shift from Process to Entity

Compare a B2-level sentence with the C2-level construction found in the text:

  • B2 (Action-oriented): Researchers used particle accelerator scans and developed AI algorithms to reduce the risks of damaging the scrolls.
  • C2 (Entity-oriented): The Vesuvius Challenge has mitigated these risks through the implementation of particle accelerator scans and the development of AI algorithms...

In the C2 version, the focus shifts from the people doing the work to the mechanisms of the work (implementation, development, mitigation). This removes the 'human' element and replaces it with 'systemic' authority.

◈ Precision through Technical Collocations

C2 mastery requires the ability to pair abstract nouns with precise, low-frequency modifiers. Observe these clusters from the article:

Structural degradation \rightarrow Not just 'breaking' or 'damage,' but a systemic loss of integrity. Technical validation \rightarrow Not just 'testing if it works,' but the formal process of proving a methodology. Virtual unwrapping \rightarrow A metaphorical compound noun that encapsulates a complex computational process into a single, digestible concept.

◈ The 'Nuanced Qualifier' Strategy

Note the use of hedging and specification to avoid overgeneralization—a hallmark of scholarly C2 English:

  • "...potentially biased summaries": The word potentially transforms a definitive claim into a scholarly hypothesis, protecting the writer from inaccuracy while still implying a critical perspective.
  • "...probable author": The shift from 'likely' to probable maintains a higher register of formality.

C2 Takeaway: To write at this level, stop asking "Who did what?" and start asking "What phenomenon is occurring?" Transform your verbs into nouns, and your adjectives into precise technical descriptors.

Vocabulary Learning

carbonized (adj.)
Converted into carbon, typically by burning or exposure to extreme heat in the absence of oxygen.
Example:The archaeologists discovered carbonized seeds that provided clues about the ancient diet.
illegible (adj.)
Not clear enough to be read; impossible to decipher.
Example:The ink had faded so significantly over the centuries that the manuscript became entirely illegible.
substrate (n.)
The underlying substance or layer on which something else (such as ink or a chemical) is applied or rests.
Example:The researcher analyzed the chemical composition of the papyrus substrate to determine its origin.
mitigated (v.)
Made less severe, serious, or painful; reduced the adverse effects of something.
Example:The use of a climate-controlled chamber mitigated the risk of further decay to the fragile scrolls.
calibrated (v.)
Adjusted precisely to a standard or a specific set of conditions to ensure accuracy.
Example:The AI model was calibrated to recognize the specific hue of ink used in the second-century texts.
Practice C2 words in a crossword