Automasi Pengekstrakan Fitur Terhadap Imej Digit Bercetak

Abstract

          This research will explain the use of Geometric Moment Invariant, Zernike Moment Invariant and Higher-Order Centralized Moment Invariant as feature extraction techniques for printed digit. It is an automation process of Geometric Moment Invariant, Zernike Moment Invariant and Higher-Order Centralized Moment Invariant for printed digit. Inter-class and intra-class analysis will be performed to show the effectiveness of feature extraction techniques using, these concepts. Based on the results, Higher-Order Centralized Moment Invariant was found to give the better result for inter-class and intra-class analysis.

Website : http://ir.fsksm.utm.my/544/