On Analysis of Invariant Characteristic for Moment Invariant Techniques

In this paper a set of equation known as invariant error computation (iec) is introduced that is used to examine directly the invariant performance properties of moment invariant techniques. the technique consists of a set of equations known as total percentage min absolute error (tpmae), percentage min absolute error 1 (pmae1), percentage min absolute error 2 (pmae2) and percentage absolute error (pae). these equations are utilized to measure the similarity between different feature vectors produced by the moment techniques studied. in order to evaluate the effectiveness of the iec, we examine the invariant properties of three moment techniques; namely; zernike moment invariant (zmi), legendre moment invariant (lmi), krawtchouk moment invariant (kmi). it is found that krawtchouk moment invariant (kmi) generated the lowest error value for the iec when compared to zmi and lmi. for instance pmae1 for kmi is the lowest with 0.1%-0.5% of error while lmi 8%-25% and zmi 8%-38% consecutively similarly with pmae2 results, kmi error is between 0.001 %-0.9% while for lmi and zmi is 0.5%-40% and 0.3%-44%. we demonstrate the effectiveness of iec in examining the invariant properties of the moment techniques.