Research On Medical Image Registration And Mosaic Based On Feature Points

Medical image registration and mosaic technology is a very distinctive research field. It’s a frontier subject in the field of medical images which will not only put all kinds of medical image information together, but also provide a new way for the clinical diagnosis.We mainly study the medical image registration based on feature points in this thesis, and probe into the key technology of medical image mosaic. Based on the analysis of various feature point extraction algorithms, we focus on the scale invariant feature transform (SIFT) algorithm and accurately extract the corresponding feature points of medical image. In order to improve the accuracy of image registration, we introduce the Mahalanobis distance affine invariance, epipolar geometry and two-dimensional projection transformation and design a gradual feature point matching algorithm.In addition, this thesis also studies the medical image mosaic algorithm based on feature points. In this thesis, we put forward an improved algorithm based on the weighted average method to ensure the quality of the image after splicing.Experimental results show that the algorithms proposed in this thesis have good validity, high accuracy and fast calculation speed.