SISTEM DETEKSI PENGGUNAAN FORMALIN PADA PANGAN IKAN ASIN MENGGUNAKAN ALGORITMA CONVOLUTIONAL NEURAL NETWORK (CNN)

Abstract

Listen

Translate article

Salted fish is a food that is very popular in the community. In addition to its delicious taste, salted fish can be purchased at an affordable price. Salted fish is fish that has gone through a preservation process using salt preservatives. The process of preserving salted fish is very dependent on natural factors such as weather and sunlight, so this preservation process cannot be used all the time. This makes salted fish producers suffer losses, so salted fish producers do everything to gain profits, one of which is using the dangerous preservative formaldehyde. The effects of consuming formalin preservatives include abdominal pain, diarrhea, nervous depression, and circulatory disorders. Based on these problems, a tool is needed that can detect the use of formalin in salted fish. In this research, one of the deep learning methods is used, namely Convolutional Neural Networks (CNN) as an image identification which is proven to be efficient in classifying images. The method is implemented by classifying formalin salted fish based on the images inputted by the user. In this study, training was carried out 3 times to find the best value. The data used in this study were 200 images of salted fish captured using a smartphone camera consisting of 100 images of formalin salted fish and 100 images of non-formalin salted fish. The results of this study are that the system can recognize objects with an accuracy rate of 92.5% with 40 times of testing which can detect 3 times identifying wrong images and 37 times correctly identifying images.