Fishes play such major roles in human life, giving important sources of food and economic sustenance across the world. However, the high incidences of diseases in fish populations significantly pose a threat to aquatic ecosystems and food security. The high mortalities seen in marine fish caught for aquariums raise several important questions regarding water quality and ecosystem health. While fish illnesses do not always show a direct link to the water quality, they are indicators for such, which simply means better testing methods are needed. While a system where the examination is comprehensive would be desirable, cost and practicality due to the complexity of such examinations restrict such activities in many instances. It is important to address fish diseases, though, as a means to maintain biodiversity and ensure sustainable fishery practices for the future.
The system's core objective is to identify diseases in fishes accurately. The machine learning model powering our system will undergo a rigorous training using a diverse dataset of fish images. There are 2 categories, healthy and infected. The model will use advanced deep-learning techniques to extract patterns from photo frames, convert them into a binary string and then train the model. The model is trained purely on clean data allowing the model to give accurate and reliable results. Together lets shape the health of our fishes together with the power of Machine Learning.
Developed by: Aathithya, Caleb and Darryan