Neural Network Classification of Corals

Science Press Release Distribution Services

Neural Network Classification of Corals

January 2, 2021 Environment 0

The recording of ecological change over time is a key challenge of contemporary ecology. The most diverse and complex of marine environments are coral reefs. In addition, coral reefs worldwide are suffering a serious decline due to the detrimental synergistic effects of global climate change, ocean acidification and seawater warming, intensified by anthropogenic eutrophication and pollution. These factors influence the decrease in live reef cover and the decrease in coral species, depending on site exposure to stressors, taking place at various rates and grades of severity. Any remediation steps include intensive monitoring, over long periods and short intervals, at numerous locations. The time-consuming, boring, manual classification of coral species and their real-time abundance in many reefs is a daunting task that is almost impossible. Deep learning (DL) has unique features for streamlining the description, analysis and monitoring of coral reefs, saving time and achieving greater reliability and accuracy compared to error-prone human results, and for managing and analysing the vast quantities of resulting data. Live coral cover and coral biodiversity are two key reef health indicators. The main objective of this analysis is to highlight the intensity and reliability of the deep learning approach to the documentation of the features of coral reefs, based on the assessment of the published application of this method to the definition of coral reefs and their assemblages of organisms. We evaluate the latest trends in the sector, describe its present constraints and future developments.

Author(s) Details

Alina Raphael
The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan 5290002, Israel.

Zvy Dubinsky
The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan 5290002, Israel.

Nathan Netanyahu
Department of Computer Science, Bar-Ilan University, Ramat-Gan 5290002, Israel.

David Iluz
The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan 5290002, Israel.

View Book :- https://bp.bookpi.org/index.php/bpi/catalog/book/358

 

Leave a Reply

Your email address will not be published. Required fields are marked *