Research On Medical X - Ray Image Mosaic And Its Improvement

With the development of medical technology, medical image mosaics technique in clinical applications has become quite common, which has great value in clinical application for image-guided orthopedic diagnosis and treatment. At present, images obtained by image detector cannot have both high-resolution and wide-field simultaneously, therefore splicing some small high-resolution images into a high-resolution wide-field image has its research value. This thesis realizes an improved method for the medical X-ray image mosaicing. The new method not only has improved the registration speed and accuracy but also enhanced the image quality.The main work of this paper is based on the SURF (Speeded-Up Robust Features) algorithm to improve X-ray medical image mosaic. Firstly, it summarizes the current scientific research of the domestic and foreign researchers in the field of image mosaicing, and then introduces the theoretical knowledge of medical image mosaicing, including the principle of image mosaicing, image registration and geometric transformation, image interpolation and similarity measure and so. Secondly, based on the SIFT (Scale Invariant Feature Transform) algorithm, the SURF algorithm is described by comparing with SIFT. In order to improve the speed and accuracy of matching and detecting feature points in the process of image registration, medical images are treated to edge detection and image enhancement before stitching, and,this paper both introduces the improves eight directions of Sobel edge detection method and the nonlinear extrapolation in the frequency space of image enhancement method. Finally, based on the original SURF algorithm and combined with the pretreatment, we skillfully finished the image mosaic. To verify the advantages of the ideas processed above, I experiment and compare with the existing algorithm in order to get an objective assessment. The results show that the improved method not only makes up many lacks of the original algorithm but also achieves a better image mosaic.