<![CDATA[Milkfish is a consumption fish, one of which is the most sold in Tarakan City. Fish is included in the food that quickly decreases the quality of its freshness. One way to see the level of freshness of fish is to look at the eyes. For people who are used to recognizing the freshness of fish, of course there will be no obstacles, but for ordinary people it will be a problem in itself. This problem can be solved by utilizing digital image processing techniques, therefore an application is needed to determine the freshness level of milkfish through fish eye images using the CNN (Convolutional Neural Network) method. The way the method works is to take the RGB value of the photo image on the eye of the milkfish, then perform a convolution by moving a 3 x 3 carnel filter on the image, so that the computer will find a new representative information, namely from the result of multiplying the part of the image with the filter used, until finally got 2 values for classification. The results of the analysis using 24 test data, where the accuracy of detection of milkfish freshness with a flashlight camera detection menu and a distance of ± 5 cm is 87.5%, detection with a camera without a flashlight and a distance of ± 5 cm is 66.67%, detection with a camera flashlight and distance of ± 10 cm is 29.17%, detection with camera without flashlight and distance of ± 10 cm is 66.67%, and detection from gallery is 91.67%. Each test uses 24 test data.]]>
Copyrights © 2023
