Handwriting Ownership Recognition using Contrast Enhancement and LBP Feature Extraction based on KNN

Every human handwriting has different characteristics. Computers can be trained to recognize handwritten ownership when they have been digitized. When handwriting is digitized into an image, the written handwriting will be in the zone measured by the unit cell. Each handwriting will have different cell sizes, so it can be used as a very useful feature in the recognition process. The Local Binary Pattern (LBP) is proposed in this study because it can extract the image according to the specified cell size. While K-Nearest Neighbor (KNN) is chosen because it has a simple algorithm by doing the calculation of the nearest distance based on the comparison of training data distance and test data, so it has a fast computation process. Contrast enhancement is also used to be proposed in preprocessing steps that aim to increase the intensity so that features are easier to extract. Thus, this study proposes a handwritten ownership recognition system with feature extraction of LBP and KNN classifier to recognize handwritten ownership. In the experimental stage used handwritten 360 images written by three different people. Where the data set consists of 300 training data and 60 test data. The test results show that the proposed method has a very high accuracy of 96.67%.