Symmetry Computation in Repetitive Images Using Minimum-Variance Partitions

In this work we propose an unified method to compute image symmetries based on founding the minimum-variance partitions of the image that better describe its repetitive nature. Then we use an statistical measure of these partitions as symmetry score. The main idea is that the same measure can be used to score the others symmetries (rotation, reflection, glide reflection). Finally, a feature vector composed by these symmetry values is used to classify the whole image according to a symmetry group. An increase in success rate, compared to other reference methods, indicate better discriminative capabilities of the proposed symmetry features. Our experimental results improves the state of the art in wallpaper classification methods.