Automatic image registration for arm radiographies

Medical image registration is a vital medical imaging application. It aligns one image to another to obtain a registered image that has the information content of both images. The main focus of this work is to use an automatic image registration of arm radiographies. Our approach is based on the following main steps: firstly, the image is binarized by thresholding technique. Secondly, the closing morphological operator is performed on the image to fill up the holes of the segmented objects. In the third step, the control points are obtained by using the skeleton algorithm. In the fourth step, the normalized cross-correlation (NCC), is adopted as the matching criterion. In the fifth step, a Ransac algorithm is applied to reject outliers. Finally, we use a geometric model estimation to map the target image onto a reference image. Experimental evaluation using our approach indicates robust and consistent image registration results.