Research On Processing And Analysis Of X-ray Image Based On Mathematical Morphology

Applying theory of digital image processing on processing and analysis of X-ray images in clinical can lighten the doctor's burden and raise diagnostic accuracy, while is advantage to realize intelligent diagnosis. Mathematical morphology is a nonlinear tool on image processing and analysis, and Objects in the images are nonlinear included relationship, so mathematical morphology is fit for image processing and analysis. According to characteristic of X-ray image, this paper presents enhancement algorithm and edge extracting algorithm of X-ray orthopedic image based on keystone, operations and property of mathematical morphology.The multi-scale fracture X-ray image enhancement algorithm based on the reconstruction focuses on the X-ray image enhancement algorithm and structuring elements selection algorithm. X-ray image enhancement focuses on denoising and enhancing the contrast. Theses operations based on reconstruction cannot bring the new edge information while enhancing images; Selecting of structuring elements includes form and scale. Choose the form of structuring elements according to experimental results of processing long bones using different form of structuring elements. And select the scales of multi-scale structuring elements according to the amount relations between different scales and the connectivity of regional in the image.Edge extracting of X-ray orthopedic image is improved in the basis of marker-controlled watershed algorithm, and mainly introduce morphological direction gradient algorithm and marker selection algorithm. According to characteristics of morphological structuring elements, the paper proposed use of the four directions of the linear structuring elements calculating direction gradient, so that the representative rate of gradient information also have directional; maker selection algorithm selects the prospects of marker under gray difference and selects the background of marker under SKIZ theory. Finally, using the markers selected modifying the directional gradient can extract the edge of bone.In addition, the paper also verify applicability of enhancement algorithm and edge detection algorithm. By calculating the statistical features of enhanced image the as histogram, proved that the proposed algorithm achieved a purpose of enhanced X-ray fracture images; By calculating the ultimate measurement accuracy, proved that extracted the edge of bone in the X-ray image using proposed algorithm can used for calculate the features and parameters.