An In-Depth Automated Approach for Fish Disease Recognition

Abstract

Fish plays a significant role in food and nutritional security in our country as well as the whole world. Owing to this reason, it becomes essential to increase the production of fish. But it is diminishing due to numerous diseases which can deteriorate the national economy. It is a fact that there is no single effective research work that has been done in regards to fish disease due to a lack of data and a high level of expertise. Consequently, our aim is to recognize the fish disease effectively that can help the remote farmers who need proper support for fish farming. Recognition of disease-attacked fish at an early stage can help us take necessary steps to prevent from spreading of the disease. In this work, we have performed an in-depth analysis of expert systems that can continue with an image captured with the help of smartphones and identifies the disease. Two set of features is selected then a segmentation algorithm is employed to detect the disease attacked portion from the disease-free portion. Furthermore, eight prominent classification algorithms are implemented accordingly to measure the performance using performance evaluation matrices. The achieved accuracy of Random forest 88.87% which is promising enough.