Pengolahan Citra untuk Identifikasi Jenis Telur Ayam Lehorn dan Omega-3 Menggunakan K-Mean Clustering dan Principal Component Analysis

Chicken eggs are divided into several types including omega-3 chicken eggs, native chicken eggs, Arab chicken eggs, and domestic chicken eggs. Visually to distinguish the type of domestic and omega-3 chicken eggs have difficulty because physically the shape of the eggshell and the color of the chicken eggs look the same. Visual inspection of the two types of chicken eggs has a weakness because it only relies on the sense of sight that has limitations, and the results are less accurate because it is very dependent on the interpretation of each consumer. This research aims to distinguish the two types of domestic and omega-3 chicken eggs which pre-processing techniques of contrast stretching, brightness, histogram equalization, changing color images to gray images, then the k-mean image segmentation process is carried out. clustering, morphological operations, dilation, and erosion. Next, the first-order statistical feature extraction is done by calculating values namely mean, variance, entropy, skewness, and kurtosis results from the histogram. The final step is to look for eigenvalues, the eigen vector PCA method used to distinguish omega-3 egg types. The results in the form of plot graphs of mean and entropy features after the second rotation show that the first-order statistical feature extraction method and PCA method can be used to significantly distinguish the types of lehorn chicken and omega-3 chicken eggs.