Computer-aided disease diagnosis in aquaculture: current state and perspectives for the future

Automation of essential processes in agriculture is becoming widespread, especially when fast action is required. However, some processes that could greatly benefit from some degree of automation have such difficult characteristics, that even small improvements pose a great challenge. This is the case of fish disease diagnosis, a problem of great economic, social and ecological interest. Difficult problems like this often require a interdisciplinary approach to be tackled properly, as multifaceted issues can greatly benefit from the inclusion of different perspectives. In this context, this paper presents the most recent advances in research subjects such as expert systems applied to fish disease diagnosis, computer vision applied to aquaculture, and image-based disease diagnosis applied to agriculture, and discusses how those advances may be combined to support future developments towards more effective diagnosis tools. The paper finishes suggesting a possible solution to increase the degree of automation of fish disease diagnosis tools.