Eggs Classification based on Egg Shell Image using K-Nearest Neighbors Classifier

Abstract

Chicken eggs are one of the foods that are widely consumed by humans. The quality of eggs will affect the nutritional quality of eggs. One method that can be used to determine the quality of the outer shell is the quality of the eggshell. This research proposes egg classification techniques based on eggshell images using the K-Nearest Neighbors (KNN) classifier based on two feature extractions, namely the extraction of Hue Saturation Value (HSV) color features, and the Gray Level Co-Occurrence Matrix (GLCM). The experiment was carried out using 100 egg images consisting of three classes, namely eggs of good quality, rotten, and defective. Of the 100 images used 21 images as testing images and the rest as training images. The test was conducted with parameter values k = 1.3, and 9 while the distance used for each k was 1.2, and 4. Based on the test results obtained the highest accuracy of 85.71%, where the parameter value k = 1; d = 2 and k = 1; d = 4.