Automated Counting of Giant River Prawn Larvae using Machine Vision

In this scholarly endeavor, we introduce a machine vision system tailored for counting giant river prawn larvae within a water-filled vessel. Our system exhibits the capacity to perform real-time and image-based counting of these organisms at a density of 500 individuals per image, achieving an accuracy exceeding 99.40%. This system offers a multitude of distinct advantages over conventional counting methodologies, including rapid processing, high precision, and reduced labor demands. Furthermore, it stands as a non-invasive, non-destructive, and cost-effective solution for counting, thus promising significant improvements in the field of giant river prawn aquaculture.