Detecting Eggs of Invasive Apple Snails on Rice Fields Using YOLOv8 Algorithm

Apple snails are invasive species that reproduce rapidly, outcompete native species, damage ecosystems, and even endanger human health. They threaten agriculture by consuming rice seedlings in their early stages and potentially transmit crop diseases, causing significant economic and environmental harm. Though various management techniques are employed to control their population, these methods primarily focus on adult apple snail invaders and cause unintended harm to non-target species. Hence, detecting them early at their embryonic stage is crucial to mitigate the negative impacts of the apple snails and their traditional management methods alike. Utilizing object detection and embedded computing boards, apple snails can be managed in an environmentally sustainable and economically feasible manner. This study introduced an object detection model that localizes apple snail egg masses, leveraging YOLOv8 - a leading one-shot object detection framework, in conjunction with the Raspberry Pi 4 Model B, creating a portable and handheld detector. The developed model achieved impressive precision, recall, m<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$A$</tex>P50, and mAP50-95 rates, respectively (92.708%, 88.6%, 94.22%, 74.62%), demonstrating their effectiveness.