Image Clustering using K-Means on Marine Products

Abstract

In this study, the researcher collected 360 marine product images consist of red snapper, prawn, silver belly, pomfret, mackerel, cuttle fish, lobster, crab and sardine to conduct try-outs at first. Secondly, images are separated from background for processing. Then from the images features are extracted via Gray Level Concurrence Matrix. Finally images are clustered according to its groupings by K-Means clustering algorithm. Since marine products are consumed by most of populaces regularly because of its health benefits, availability of nutrients and low cost. For that reason all and sundry can buying. This research helps to identify them by their physical appearances. Marine products have eye-catching altered physiognomies which are cherished to extricate and conclude a specific category. These physical appearances comprise of size, shape, texture, and color. This research succeeds 83% accuracy for bunch the images into nine clusters.