# A Deep Learning-Based Method for Non-Destructive Estimation of Carbonate Carbon Storage in Biogenic Shells on Marine Engineering Materials

## TL;DR

A non-destructive deep learning method is developed to estimate carbonate carbon storage in marine biofouling shells using in situ images.

## Contribution

An improved Mask R-CNN framework enables automated, non-destructive quantification of shell carbonate carbon storage on marine surfaces.

## Key findings

- Image-derived shell dimensions strongly correlate with manual measurements (R2 = 0.95).
- Panel-scale carbonate carbon storage is estimated with errors below 15%.
- Peak carbonate carbon storage on uncoated PVC panels reached ~1.061 g per panel in September.

## Abstract

What are the main findings?
A non-destructive framework is developed to quantify shell carbonate carbon storage on marine engineering surfaces.An improved Mask Region-based Convolutional Neural Network (Mask R-CNN) enables automated identification and shell dimension extraction of barnacles and bivalves from in situ images.Image-derived shell dimensions show strong agreement with manual measurements (R2 = 0.95).

A non-destructive framework is developed to quantify shell carbonate carbon storage on marine engineering surfaces.

An improved Mask Region-based Convolutional Neural Network (Mask R-CNN) enables automated identification and shell dimension extraction of barnacles and bivalves from in situ images.

Image-derived shell dimensions show strong agreement with manual measurements (R2 = 0.95).

What are the implications of the main findings?
Panel-scale carbonate carbon storage is estimated with errors below 15% under complex nearshore conditions.The proposed framework provides a non-destructive approach for comparative analysis and long-term monitoring of biofouling on different surface materials.Enables comparative quantification of biofouling and carbon storage across different materials.

Panel-scale carbonate carbon storage is estimated with errors below 15% under complex nearshore conditions.

The proposed framework provides a non-destructive approach for comparative analysis and long-term monitoring of biofouling on different surface materials.

Enables comparative quantification of biofouling and carbon storage across different materials.

Hard-shelled organisms colonizing marine engineering surfaces accumulate carbonate inorganic carbon in their shells, yet quantification typically relies on destructive sampling, hindering long-term monitoring. This study develops a deep learning-based, non-destructive framework to estimate shell carbonate carbon storage from in situ images. Panels of different surface materials were deployed in the nearshore waters of Liuheng Island (Zhoushan) and monitored for five months, yielding 90 panel images from June to October. An improved Mask R-CNN identified barnacles and bivalves and extracted shell dimensions, which were combined with allometric relationships and measured shell carbonate carbon fractions (12.07% for barnacles; 12.14% for bivalves) to estimate carbon storage. Peak colonization occurred on uncoated polyvinyl chloride (PVC) panels in September (~110 individuals per panel), corresponding to 1.061 g carbonate carbon per panel. The model achieved recall/precision of 0.86/0.89 under complex nearshore conditions; image-derived dimensions agreed with manual measurements (R2 = 0.95). Allometric models showed R2 of 0.82 (barnacles) and 0.90 (bivalves), and panel-scale estimation errors were <15%. The method enables non-destructive quantitative characterization and comparison of shell carbonate carbon storage across materials and exposure conditions for long-term monitoring.