Document Image Registration Using Geometric Invariance and Hausdorff Distance

A new method of document image registration based on geometric invariance and Hausdorff distance is presented in this paper. Firstly, geometric invariance is used to obtain candidate feature points. Then Hausdorff distance is used to determine the final fine transformation between the input document image and the reference document image through the corresponding points set. Experiments indicate our method leads to accurate and robust registration, and can extract the filled-in information reliably.