Using mobile-based augmented reality and object detection for real-time Abalone growth monitoring

Abalone are becoming increasingly popular for human consumption. Whilst their popularity has risen, measuring the number and size distribution of Abalone at various stages of growth in existing farms remains a significant challenge. Current Abalone stock management techniques rely on manual inspection which is time consuming, causes stress to the animal, and results in mediocre data quality. To rectify this, we propose a novel mobile-based tool which combines object detection and augmented reality for the real-time counting and measuring of Abalone, that is both network and location independent. We applied our portable handset tool to both measure and count Abalone at various growth stages, and performed extended measuring evaluation to assess the robustness of our proposed approach. Our experimental results revealed that the proposed tool greatly outperforms traditional approaches and was able to successfully count up to 15 Abalone at various life stages with above 95% accuracy, as well as significantly decrease the time taken to measure Abalone while still maintaining an accuracy within a maximum error range of 2.5% of the Abalone's actual size.

Automated summary

Abalone are increasingly popular for consumption, but measuring their number and size distribution in existing farms is challenging. Current methods rely on manual inspection, which is time-consuming and results in mediocre data quality. To address this, we propose a mobile-based tool that combines object detection and augmented reality for real-time counting and measuring of Abalone. Our experimental results showed that the proposed tool outperforms traditional approaches, achieving above 95% accuracy in counting Abalone and reducing measurement time, while maintaining an accuracy within a maximum error range of 2.5% of the Abalone's actual size.