Prospecting for Australia's Submerged Landscapes through Machine Learning

通过机器学习勘探澳大利亚的水下景观

• 批准号: DE220100550
• 负责人: Dr John McCarthy
• 金额: $ 31.86万
• 依托单位: Flinders University
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Early Career Researcher Award
• 财政年份: 2022
• 资助国家: 澳大利亚
• 起止时间: 2022-01-01 至 2024-12-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DE220100550
• 关键词: Prospecting; Australia; Submerged; Landscapes; through

This project aims to apply machine learning to image-based seabed surveys to prospect for submerged Aboriginal archaeological sites, beginning with both of Australia’s only known sites. This will be the first attempt globally to develop a technique to pinpoint potential archaeological material within large area surveys, tagging features which can then be tested through scientific diving. Expected outcomes are workflows that pair machine learning algorithms, marine robotics and scientific diving to greatly enhance prospecting efficiency. This will enhance knowledge and benefit management of these resources on Australia’s continental shelf and beyond, reducing the impacts of offshore industry on cultural heritage.

该项目旨在将机器学习应用于基于图像的海底调查，以前景淹没的土著考古遗址，从澳大利亚仅有的两个已知遗址开始。这将是全球首次尝试开发一种技术，在大面积调查中查明潜在的考古材料，标记特征，然后通过科学潜水进行测试。预期成果是将机器学习算法、海洋机器人和科学潜水配对的工作流程，以大大提高勘探效率。这将加强对澳大利亚大陆架及以外地区这些资源的了解和利益管理，减少近海工业对文化遗产的影响。