Indonesia is a nation rich in a variety of cultures ranging from customs, culture, livelihoods, social, even to culinary has become a specialty of this nation. One of them is pempek or commonly called empek-empek. A snack originating from the Palembang area, South Sumatra is very popular among Indonesians from the lower class, middle class, to the elite who are usually used as a side dish. Pempek is made from softly ground fish meat mixed with starch or sago flour, and with the addition of other ingredients including eggs, garlic, flavoring and salt. Currently, many household industries (IRT) in processing fish meat are still not in accordance with the authenticity of Palembang so that the creation and taste of pempek from one industry to another will be different. In this research, KNN (K-Nearest Neighbors) algorithm is juxtaposed with GLCM (Gray Level Co-accurence Matrix) algorithm in processing the image of pempeki dough. By using GLCM and KNN algorithms, this research aims to identify the processed dough that matches the original Palembang pempek snacks where the taste and texture are like the original. The GLCM algorithm is used to extract patterns from the image of pempek dough. In this study, the accuracy obtained from the KNN algorithm is 83% by using the K value of K = 3, K = 5, K = 9, K = 11.
Copyrights © 2026