Automating Lobster Length Assessments Towards Facilitating Rapid Fisheries-Assisted Sampling

The sustainable management of fisheries relies heavily on accurate and timely data collection, yet small-scale fisheries often struggle with the limitations of time and resources. This study presents an approach to automating lobster length assessments using a fisheries-assisted sampling technology (FAST) system. The system integrates a fixed-size baseboard, a consumer-grade smartphone or tablet camera, and a Bluetooth-connected scale, enabling rapid and precise data collection. Key-point detection, facilitated by the DeepLabCut framework, and a deterministic algorithm for target region detection are used to accurately measure lobster carapace lengths from digital images. This method significantly reduces manual measurement time and error, allowing for large-scale data collection. The system's effectiveness was validated through field tests, showing high accuracy in identifying key-points on the lobster carapace. This approach not only enhances the efficiency and accuracy of data collection but also has the potential to be adapted for various marine geographical regions, contributing to better-informed fisheries management decisions. Future work will focus on integrating the system into broader fisheries data collection frameworks.