Automated Visual Quality Detection for Tilapia using MobilenetV2 Convolutional Neural Network

As the aquaculture industry expands, it becomes increasingly crucial to prioritize the quality of seafood products. This research presents a novel method for automated visual quality detection of Tilapia utilizing the MobileNetV2 Convolutional Neural Network. By taking advantage of the capabilities of deep learning techniques, the researchers trained an AI model via multiple layers of diverse dataset that underwent many data augmentation processes to provide more accurate results. MobileNetV2, renowned for its efficacy in image classification tasks, helped the model to precisely evaluate the fish characteristics such as its eyes and body, in the end, the study managed to get over 73% accuracy and a range of 0.5 to 0.8 F1 score in both tests which is a good result. The proposed automated system not only streamlines the quality control process of Tilapia in the fish industry but also holds promise for enhancing overall efficiency and reliability in overall fish production. Extensive experiments by the researchers validate the model's impressive performance in achieving accurate detection results, marking a significant advancement in quality assurance within the aquaculture sector.