Table 7. Results of the hkaAICnorm MATLAB script on the pattern with a large amount of added noise that underlies Fig. 6 ▸ for geometric model selection by G-AIC value minimization using inequality (9b) .

Plane symmetry group to which the image data have been symmetrized	Sum of squared residuals of complex Fourier coefficients	Sum of squared residuals of Fourier coefficient amplitudes	No. of Fourier coefficients in the geometric model of the image data
p2	0.0061	none	275
p1m1	1.5353	0.0039	271
p11m	1.5320	0.0039	271
p1g1	0.0069	0.0039	265
p11g	0.0078	0.0039	270
c1m1	0.0085	0.0041	269
c11m	0.0074	0.0041	269
p3	1.7565	1.2029	306
p2gg	0.0098	0.0039	264
c2mm	0.0115	0.0041	269
p4	0.0088	0.0028	276
p4mm	1.5876	0.0053	276
p4gm	0.0109	0.0051	266