Fish Freshness Classification Using Combined Deep Learning Model
Muhammad Abu Rayan, Abdur Rahim, Md Abir Rahman, Md. Abu Marjan, U. A. Md. Ehsan Ali
Abstract
Fish is a great source of protein. It also contributes a lot on the economy of Bangladesh. Besides fulfilling our needs, we also export fish in different countries. So maintaining fish freshness is an important issue. This paper proposed an automatic method for classifying fish freshness based on combined deep learning model using image processing. The process extract features using VGG-16 neural network architecture and Bi-directional Long Short Term Memory build ML (Machine Learning) model. Here we used Nile Tilapia as our sample fish. This combined deep learning model shows an incremental leap in finding fish freshness. The proposed model has achieved 98% accuracy on testing dataset.