Classification Types of Wood Furnitures Using Gray Level Co-Occurrence Matrix and K-Nearest Neighbor

In this study, the author wants to test the maximum accuracy obtained by GLCM feature extraction with datasets of 4 wood classes, namely teak, mahogany, mango, and sungkai wood, combined with the K-Nearest Neighbor (K-NN) classification. I was using four features in GLCM, Homogeneity, Contras, Correlation, and Energy. Some journals think that GLCM can calculate the pixel surface well in a still image. For this reason, I want to examine more deeply whether GLCM can classify well on the surface of a wooden furniture structure. As a reference to determine the accuracy value, I use K-NN euclidean distance KNN has the power of multiclass classification. This study succeeded in obtaining the lowest accuracy value at 50% and the highest value at 77.5%.