Extraction Of Lumbar Vertebral In Medical X-ray Image

With the development of computer aided medical technology,medical X-ray become a hot research topic in the field of biomedical image analysis gradually.As the interested region in the lumbar vertebral X-ray image,it is important to extract vertebral bodies from lumbar vertebral image,which has great reference value for the diagnosis of lumbar disease.However,due to the original existing interference factors in the X-ray image,such as low contrast,edge blur and noise pollution,it faces numerous challenges in the analysis of vertebral body information in lumbar vertebral image.The thesis gives the research on interest region extraction in lumbar vertebral X-ray image,from the view of X-ray image characteristics and medical image processing,it analyzes the deficiency of image processing algorithms in medical X-ray image processing aspects thoroughly.And then,it proposes the sharpening mask enhancement method with the idea of improved haze removal algorithm,the comprehensive edge detection method based on phase consistency and the interested region extraction method in medical X-ray image on the basis of generalized Hough transform.The main work of this thesis is as follows:Firstly,the imaging principle of X-ray image is analyzed.On the basis of feature of X-ray image,it carries out the enhancement processing of medical X-ray image,and studies the image enhancement method in spatial and frequency domain.Meanwhile,it gives the research on histogram equalization in spatial domain,adaptive histogram equalization with contrast limited,smooth filtering,sharpening filter,gray level transform algorithm,as well as Fourier transform,high-pass filtering,low-pass filtering in the frequency domain.And then,according to the characteristics of fog haze image and medical X-ray images,it introduces guided image filtering method with haze removal theory.It designs the framework of image preprocessing and keeps the image edge features well.Secondly,the image edge extraction algorithm is studied.It researches the Sobel operator,Roberts operator,Prewitt operator,log operator and canny operator through experiment carefully.And then,it realizes the comprehensive edge detection method based on the phase consistency,which is used as the edge detection scheme of generalized Hough transform.Finally,it gives the determination and division of the interested region.It studies the extraction method of interested region,and implements the interested region extraction of X-ray medical image with generalized Hough transform.The key of the success of generalized hough transform is the segmentation of edge information with matching.Through the test of large number of X-ray lumbar image,it implements the effective segmentation.The thesis combines the haze image with medical X-ray image processing technology,with the enhancement of medical X-ray image,it keeps the image edge details.With the process of region growing on extracted edge,it can reveal the whole vertebral body and minimize noise.The customized lumbar template can extract the coarse-grained and fine-grained of interested region in lumbar vertebral X-ray image effectively,and provide more reliable basis of the diagnosis of disease for doctors.