Early Detection System Based on Artificial Intelligence as an Innovative Solution for Vaname Fry Farming

Abstract

Indonesia has great potential in the fisheries sector, especially in vaname shrimp farming which plays an important role in fisheries exports. One of the main challenges in shrimps farming is the availability of quality fry. Cuurently, the process of counting fry is done manually, which is less efficient and accurate. To improve efficiency, researchers developed digital image technology in the form of an artificial intelligence (AI)-supported application that is a solution in the automation of fry counting, health detection, and early detection of disease. The method used involves Convolutional Neural Network (CNN) for image analysis and Long Short-Term Memory (LSTM) for sequential data processing. The result showed that the integration of CNN and LSTM was able to classify the condition of the fry with 85% accuracy in less than 5 minutes, compared to the manual method which took longer. The developed ‘VannameiAI’ application facilitates farmers in counting the number of fry, detecting the health of fry according to the fry standard before stocking, and storing harvest data. Thus, this technology has the potential to increase the efficiency and productivity of shrimp farming in Indonesia.