Fish identification and freshness classification through image processing using artificial neural network

Abstract

The demand for fish is continuously rising due to its high nutritional value. Inexpert manual determination of fish freshness can cause false assessment and result to the possibility of food poisoning. This study developed an android application that automatically identifies the three most consumed fish in the Philippines, namely milkfish, round scad, and tilapia. Through image processing, the application classifies the freshness of the fish from level 1 (stale) to level 5 (fresh) by using the RGB values of the eyes and gills as well as determining its remaining shelf life. The software was developed by iterative learning of a feed forward neural network with 30 fish samples per species that were used to obtain a total of 800 images each for the eyes and gills. The results of the study showed that the device yields acceptable results in identifying the fish and in determining its freshness.