Research On Portable Medical X-ray Image Mosaic Technology

Image mosaic is one of the important research area of computer vision,image processing and computer graphics. Large fields of view and high-resolution image can used to construct the application in virtual reality,medical image processing,remote sensing and so on.At present,X-ray perspective equipment commonly used in clinical orthopaedics to provide only limited field of vision image. So doctors need to diagnosis and treat the process of image reconstruction from a series of a complete panoramic images, some overlap images have been seamlessly splicing to create ultra-wide viewing image.The paper summarize the effort of medical image mosaic that the researchers have achieved. The thesis gives the steps and classification of image mosaic and the appropriate algorithm and the simulation of the entire process in portable x-ray radiography.This paper studies the geometric correction of splicing images, image preprocessing including histogram equalization,edge sharpening,regional segmentation,contour extraction and image registration.Focused on the image registration method. Based on the template matching method, the ratio of matching ,moravac feature extraction and harris corner detection algorithm to match images. When we use harris corner detection algorithm.Firstly,we analyze various approach to extract feature points—corner. Then,we purify the inliers which are extracted by Harris Detector using a robust filter strategy and insure the selection after purifying estimated the transform relationship precisely.We decrease the constraint of extracting feature points and match nicely.So the images still can be matched pericely from those obtained in the flexible condition. Then,we do a detailed analysis and summary to the experimental results. Finally,an operator is used to complete the smooth image mosaic and display,and the computer program of the work are completed.