Penerapan Metode Convolutional Neural Network Dalam Klasifikasi Kesegaran Ikan Mungkus Berdasarkan Citra Mata dan Insang Ikan

Abstract

Mungkus fish (Sicyopterus stimpsoni) is a type of freshwater fish that is a typical mascot of Kaur Regency, Bengkulu Province. This fish lives in clear, fast-flowing waters, and is known for its ability to stick to rocks using a special structure on its stomach called cupak. Mungkus fish has high economic value and is consumed daily by the local community. However, high demand is not balanced with adequate availability, resulting in increasingly expensive prices. In addition, the lack of public knowledge regarding the assessment of fish freshness causes the risk of consuming fish that is not fresh, which has the potential to endanger health. Traditional assessment of fish freshness based on physical parameters such as eyes, gills, and meat texture is considered less accurate and requires special expertise. Therefore, this study proposes the use of Convolutional Neural Network (CNN) to classify the freshness of mungkus fish based on eye and gill images. CNN is able to extract complex features from images without the need for manual extraction. The application of this method is expected to provide an objective, efficient, and accurate solution in assessing the freshness of mungkus fish, as well as being beneficial for fishermen and consumers.