Identifying Prawn Disease Using Improved CNN

Prawns are a vital global food source, but their increased production faces challenges due to prawn diseases causing significant economic losses for fish farmers. Detecting diseases promptly is essential to protect the aquatic ecosystem and human health. However, India’s current manual detection method often yields inaccurate results, necessitating a quick and cost-effective alternative. This research utilized Convolutional Neural Networks (CNNs) and transfer learning to address the issue, demonstrating the effectiveness of these techniques in various computer vision applications. The study specifically targeted classifying three prawn diseases: Yellow Head Virus (YHV), White Spot Prawn Disease (WSP), and Black Gill Disease (BG). Additionally, the study focuses on identifying diseased prawns from a pool of fresh prawns. By using advanced techniques such as Data Augmentation and Transfer Learning with different CNN models, including AlexNet, VGG16, ResNet50, and MobileNet, the research achieved an impressive 92% accuracy in classifying the diseases, with ResNet50 proving to be the most effective. This outcome highlights the significance of ResNet50 for prawn disease classification, providing valuable insights for disease management and safeguarding prawn farming and aquaculture practices.