期刊
MARINE ECOLOGY PROGRESS SERIES
卷 598, 期 -, 页码 213-231出版社
INTER-RESEARCH
DOI: 10.3354/meps12531
关键词
Otolith; Computer-assisted age and growth estimation; CAAGE; Image analysis; Computational model
资金
- Cefas
- University of East Anglia (UEA)
- Directorate For Geosciences
- Division Of Earth Sciences [1561622] Funding Source: National Science Foundation
Otoliths or ear-stones are hard, calcium carbonate structures located within the inner ear of bony fishes. Counts of rings and measurements of seasonal growth increments from otoliths are important metrics for assessment and management of fish stocks, and the preparation and microscopic analysis of otoliths forms an essential part of the routine work undertaken by fisheries scientists worldwide. Otolith analysis is a skilled task requiring accuracy and precision, but it is laborious, time-consuming to perform, and represents a significant cost to fisheries management. In the last 2 decades, several attempts to apply 'computer vision' (systems that perform high-level tasks and exhibit intelligent behaviour) in otolith analysis have been reported. Although considerable progress has been made and several prototype systems developed, laboratories have been reluctant to adopt image-based computer-assisted age and growth estimation (CAAGE) systems. This paper surveys applications of CAAGE, focusing on their utility for automated ageing using images of otolith macrostructure. A cost-benefit analysis of CAAGE of cod, plaice and anchovy shows that computer vision performs relatively poorly compared with morphometric techniques. However, there is evidence that information from visual features can boost the performance of morphometric CAAGE, and further work is needed to develop effective frameworks for this integrated approach. The cost benefit of these systems might be attractive to smaller laboratories that are already using age-length keys derived from otolith morphometrics for management of smaller artisanal fisheries.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据