A new method for image normalization

Image normalization is very useful in image understanding systems. In general, there are four basic forms of distortion of planar patterns: translation, rotation, scaling and skew. The authors propose a novel method which basically solves all problems in image normalization. First, we compute the covariance matrix of a given pattern. Then, we rotate the pattern according to the eigenvectors of the covariance matrix, and scale the pattern along the two eigenvectors according to the eigenvalues to bring the pattern to its most compact form. The processed pattern is invariant to translation, scaling and skew. Only the rotation problem remains unsolved. The final step is to rotate the image with respect to the image ellipse tilt angle, which is determined by the three second-order central moments, to make the pattern invariant to rotation. Thus, the resulting pattern is invariant to translation, rotation, scaling and skew. The correctness and effectiveness of the proposed method are approved by numerical experiments on aircraft images.