MR V NIVEDAN,DR K POORNIMA
DOI: https://doi.org/Shrimp farming around the world is significantly threatened by the White Spot Syndrome Virus (WSSV), which causes enormous mortality and financial losses owing to inaccurate or delayed diagnosis. With the use of static photos, this study provides CrustaScope, an AI-powered predictive visual detection system appropriate for the early diagnosis of WSSV in prawns. To increase feature discrimination, a Squeeze-and-Excitation (SE) attention block is added to the MobileNetV2 backbone, which has been fine-tuned for the model. Transfer learning is utilised to train a binary classifier with swish-activated thick layers on a bigger dataset of photos of both healthy and unwell prawns. With a clear separation between specimens that tested positive for WSSV and those that were healthy, the model displayed good classification performance. CrustaScope, which is intended for offline usage on low-resource local PCs, gives prawn producers a valuable and non-invasive approach to assist fast diagnosis and treatment. It is particularly well-suited for small-scale and rural aquaculture operations thanks of its lightweight design, limited hardware needs, and user-friendliness.
