Fish Disease Detection to Sustain Hatchery and Pond Production System

Abstract

Diagnosis of fish disease in aquaculture is a necessary process and needs an exceptionally high level of competency to sustain hatchery and pond production systems. Developing an system to overcome the challenges faced by fish farmers in stopping the spreading the disease that leads to economic loss, is a crucial task. A crucial initial phase in preventing the spread of disease is early identification of diseases in fish. The fish disease usually propagates quickly through the water, affecting large numbers of fish and causing financial loss to the farmers. Since tilapia aquaculture is one of the methods for producing food that is expanding the quickest and has the highest export value, we’d like to know more about the fish disease that affects this sector. The research uses the pathogen-infected fish. System is developed by working perfect image processing and machine learning techniques together. The proposed work has two phase. Image pre-processing has been used in the first phase to, respectively, reduce distortion and magnify the image. In the second section, the system extracts the relevant information and uses machine learning approaches for recognising the diseases. A trained machine learning model has been deployed to the first fraction's processed images. Then, using the chosen fish image dataset to study the fish disease, the research integrates an extensive experiment combining different methods.