This paper presents an effectual method for "Online signature intrusion system using KNN algorithm". Increase of the digitalization in the world makes a growth of web applications in personal authentication at different fields. To obtain more accuracy rate than static feature-based signature recognition systems we performed several experiments on most widely recognized data sets like MCYT-100 and SUSIG datasets to detect both skilled and random forgeries. The verification of signature is accomplished on the feature extraction of an image from the signature using K-Nearest Neighbor(K-NN) and Gabor filter which gives a classification of the test signature as genuine or not. Depend upon the Pre-processing of the image with the existed database values it gives the result as authorized or not.
Volume 12 | Issue 2
Pages: 1072-1079
DOI: 10.5373/JARDCS/V12I2/S20201137