Comparison Between Zernike Moment and Central Moments for Matching Problem

Abstract

Moment invariants have wide applications in image recognition since they were proposed. The recognition problem is very often connected with image reconstruction technique to determine a desired set of invariants for use feature extractor in the recognition system. The low order moments are found to be related to the global properties of the image, while high order moments contains information about details. It was found that the low order moments are not efficient for image recognition because the general shape of different objects can be similar and do not allow distinguishing one object from another. For this reason, some preprocessing have to be done to enhance this weakness , and use low order moments to overcome the weakness and test it for image recognition, image transformation (rotation, flip, translation and scaling). The main difficulty in the application of the invariant moments is an absence of the theoretical methods of estimation their efficiency in recognition tasks. In this paper, the goal is to analyze the significance of Zenike moments and central moments of different orders and compare them to determine who the best for matching in cases of is Image transformation. The accuracy for matching is almost (90%) for Zerinke moments and (95%) for central moment.