Indonesia boasts a diverse cultural heritage, one of which is the regional languages that possess unique scripts across the archipelago. An example of this is the Lontara script, used by the Bugis and Makassar communities. The Lontara script is one of the scripts in Indonesia that is endangered due to the passage of time. The K-Nearest Neighbors (K-NN) algorithm can be a tool used to recognize patterns in the Lontara script. The principle of K-NN is quite simple, namely matching the similarity of new data with the nearest test data. In this research, K-NN is used for classification and Histogram of Oriented Gradients (HOG) for feature extraction. Based on the research conducted using a testing scheme with an image size of 32×32 pixels, a dataset split of 90:10, and a k-value of 5, an accuracy result of 0.8525 was obtained.
Copyrights © 2024