Research Of The Medical Image Processing Algriothm Based On Fuzzy Distance Transform

With the emergence and development of the digital imaging technology, medical imaged have widely used in clinics and diagnosis. CT image has become a important method in clinics for its high resolution and little harm to human beings. However, the complexity of human tissue and the limitations of imaging system lead to the CT image is fuzzy and have a poor contrast, these factors will directly affect the accuracy of clinical diagnosis results. In order to obtain a correct diagnosis, provided a effective method for doctor to analyze the small structure and the lesions. It has a very important application value for computer diagnostic system and clinical diagnosis to make a deeply proceeding on the medical image. In this thesis, we first make the analysis and study of the characteristics of medical CT images and the theory of the fuzzy distance transform, and then in-depth study the technology of pruning skeleton method and the method of pulmonary fissure segmentation. Aim at the appearance of burr in traditional skeleton extraction algorithm and the difficulty of extracting pulmonary fissure, we bring some image processing method corresponding it. The main achievements are summarized as follows:(1)In-depth studying and researching the theory of fuzzy distance transform, analyzing the advantages and stability of the fuzzy distance transform in the fuzzy image processing by comparing the distance transform and fuzzy distance transform. And achieving the method use the dynamic programming methods(2) We have proposed a skeleton pruning algorithm based on fuzzy distance transform. This paper presents a skeleton pruning algorithm which can effectively remove the burr and fully reflect the morphological characteristics of the object. The theoretical basis of the algorithm is fuzzy distance transform, computing the fuzzy distance transform values of each tail in the crude skeleton, achieving the skeleton pruning according to the dynamic threshold determined by the characteristics of the skeleton to judge whether the skeleton branches should be removed and using the classification pruning method . The experiment results show that the algorithm can ensure the accurately and continuous of the skeleton during the pruning operation.(3) We have proposed a pulmonary fissure extracting method based on tracking the maximal neighborhood point. This method make the fuzzy distance transform on the contour image of the pulmonary fissure. Then the key points of the pulmonary fissure can be received by combine the local maximum of the FDT image and the local minimum of the gradient image and the trend of pulmonary fissure is ensured. Finally, a continuous pulmonary fissure can be received by tracking the maximal neighborhood point. The experiment shows that the method can efficiently extract the pulmonary fissure.