Mobile-based Fish Quality Detection System Using K-Nearest Neighbors Method

Mujair fish (Oreochromis Mossambicus) is a popular consumption fish in Indonesia. The fish is a freshwater fish found in rivers, ponds, and lakes, with a salinity of less than 0.05% for the breed. Mujair fish is widely consumed by the public as a cheap and tasty fish that is often found in traditional markets or modern markets. The fish is often sold in a fresh condition (fresh) as well as through the process of freezing (frozen). However, consumers sometimes do not fully know the information about the fish condition. Too long storage process causes the physical changes of the fish into blurry eyes, colors tend to fade, soft fish meat texture, and unpleasant smell. In this research, the process to visualize fresh fish images (suitable for consumption) and not fresh (not suitable for consumption) can be detected using K-Nearest Neighbor (K-NN) using a Smartphone. The purpose of this study is to determine the results of the calculation accuracy using the K-Nearest Neighbor (K-NN) Algorithm with the method of digital image processing of Mujair fish, so consumers can choose whether the fish is suitable for consumption or not, using a smartphone as a visualization that also can be implemented as an additional facility in an online shop.