<![CDATA[The pace of technological development introduces the automatic identification of signature authenticity which is an important task in many activities requiring legitimate evidence. The process of authenticating signatures begins with preprocessing, which consists of gray transformation, median filter, binary transformation, cropping, and edge detection. After the process of preprocessing followed by the process of determining the extraction of form characteristics with the method of Shape Feature Extraction Techniques consisting of area, perimeter, centroid, rectangularity, eccentricity, roundness. Then classified based on training data obtained from calculations Shape Feature Extraction Techniques. After classification with K Nearest Neighbor then done calculation process Mean Average Precision to determine the authenticity of signature and percentage calculation of Mean Average Precision. In the system accuracy test results obtained 61% accuracy with the retrieval of random data for 25 data. Then obtained an accuracy of 61% accuracy with the retrieval of random data for 15 data and 58% on the retrieval of 5 data. Highest accuracy was obtained on the largest data collection with an accuracy of 61%.]]>
Copyrights © 2018
