Scholarlink

<p>To assess the potential ecological effects of marine renewable energy (MRE) devices, it is essential to establish a baseline of ecosystem conditions prior to disturbance and to monitor throughout the installation, operation, maintenance, and decommissioning phases. The design of mitigation and compensation measures depends on the difference between baseline conditions and the impacted ecosystem. In this study, we develop and evaluate a semi-automated methodological proof-of-concept using computer vision to assess environmental changes around MRE devices. Using fish monitoring as a case study, we reduced image processing time by eliminating empty frames and detecting fish in the remaining frames. Species identification was carried out by experts using a public database for comparative analysis. We also conducted an identification exercise for a single fish species and tested a monocular depth-estimation method (3D reconstruction from 2D images) to measure distances between organisms and devices. Finally, we propose a potential methodological framework for integrating computer vision across the life cycle of MRE devices to monitor both natural and anthropogenic processes, including space colonization, trophic interactions, pollution, and the effects of underwater structures. The merits and drawbacks of the approach are discussed.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A computer vision-based approach to monitor changes in ecosystems associated with marine renewable energy projects

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月28日

查看全文
  • Alejandra Alamillo-Paredes,
  • Elio Guarionex Lagunes-Díaz,
  • Octavio Pérez-Maqueo,
  • M. Luisa Martínez,
  • Andrea Copping,
  • Edgar Mendoza,
  • Rodolfo Silva,
  • Norman Mercado-Silva

摘要

To assess the potential ecological effects of marine renewable energy (MRE) devices, it is essential to establish a baseline of ecosystem conditions prior to disturbance and to monitor throughout the installation, operation, maintenance, and decommissioning phases. The design of mitigation and compensation measures depends on the difference between baseline conditions and the impacted ecosystem. In this study, we develop and evaluate a semi-automated methodological proof-of-concept using computer vision to assess environmental changes around MRE devices. Using fish monitoring as a case study, we reduced image processing time by eliminating empty frames and detecting fish in the remaining frames. Species identification was carried out by experts using a public database for comparative analysis. We also conducted an identification exercise for a single fish species and tested a monocular depth-estimation method (3D reconstruction from 2D images) to measure distances between organisms and devices. Finally, we propose a potential methodological framework for integrating computer vision across the life cycle of MRE devices to monitor both natural and anthropogenic processes, including space colonization, trophic interactions, pollution, and the effects of underwater structures. The merits and drawbacks of the approach are discussed.