Silk Shield: Empowering Sericulture with AI-Driven Disease Detection

This study conducts a thorough exploration of the automated identification and categorization of diseases in silkworms, focusing particularly on four prevalent conditions: flacheria, grasseria, mascardine, and pebrin, alongside a group representing unaffected specimens. Employing the EfficientNet algorithm, an innovative methodology is crafted to precisely detect and classify silkworm diseases based on image data. This approach involves systematically collecting silkworm images displaying symptoms associated with each disease, followed by meticulous preprocessing and extraction of features. Through the utilization of transfer learning with the EfficientNet architecture, the model parameters are fine-tuned to optimize its performance in disease classification. The efficacy of this method is assessed using rigorous experimental procedures and standard metrics, showcasing its effectiveness in distinguishing between diseased and healthy silkworms with remarkable accuracy and reliability. These findings offer significant insights into the field of sericulture by presenting a robust approach for disease identification and management within silkworm populations, potentially enhancing productivity and sustainability in sericulture industries globally.