00000ctm a22000004a 4500 UP-99796217603116368 Buklod 20230503092542.0 m |o d | ta 070718s2002 xx r |||| u|eng d (iLib)UPD-00000179812 DENGII eng LG 993.5 2002 C65 B35 Balanag, Roschelle C. KRoMe signature verification Roschelle C. Balanag, Kenneth Clarke M. Cruz, Anna Meliza R. Tinio. 2002. vii, 26 leaves ill. Thesis (B.S. Computer Science)--University of the Philippines Diliman. The KRoMe Signature Verification Software is a tool that recognizes a signature based on existing samples or templates. The software is trained using a back propagation neural network. The processed signature image serves as the input to the neural network. Other pertinent statistics are also displayed to provide the user with data from which the software's conclusion is drawn. We have produced a software that recognizes a person's signature to some degree of accuracy. If the signature is trained, a 90 percent accuracy is produced, whereas if the signature is not trained, the degree of accuracy varies. It may range from 40 to 80 percent. Hence, this software would not only be useful as a stand-alone program, it can be integrated into banking systems, credit card company systems, and even in criminal investigation systems. Fingerprints--Identification--Automation. Automatic fingerprint identification system (AFIS). KRoMe (Computer program). Cruz, Kenneth Clarke M. Tinio, Anna Meliza R. Thesis FI UP UPD DENG-II LG 993.5 2002 C65 B35 Thesis