Signature Verification Using The K-Nearest Neighbor (KNN) Algorithm and Using the Harris Corner Detector Feature Extraction Method

Abstract

The security of the transaction process is very important in this day and age. Signatures can be used as a means of guaranteeing the security of a transaction other than fingerprints. However, the threat of signature forgery for those who use signatures as security is still very high and frequent. In this research, we will verify the authenticity of a signature and test it using the K-Nearest Neighbour (KNN) algorithm and the Harris Corner feature extraction method. There are two kinds of distance calculations that will be used in the K-NN algorithm, namely by calculating the distance from Euclidean Distance and Manhattan Distance. The k value at KNN taken is at k = 1, k = 3, and k = 5.