Segmentation Algorithm Based On Wavelet Transform, And Gradient Vector Flow Image

The development of medical technology makes us observe straightly the real components of human inherent mechanics. We can gain the pictures of human inherent organization through medical imaging technology, like CT, MRI, X-ray, and so on. After that it is helpful for doctors to diagnose the state of an illness with medical image processing which contains edge-detecting, matching, moving track and body token. However, the computing image division is one of the most important fields in medical image processing. The traditional methods of image processing are not suitable for processing the medical images because of the diversity and different quality of them. In this thesis, we apply the wavelet transform and degree-vector model to divide and process the images of gastric ulcer. Our aims are to detect the contour of gastric ulcer and realize semiautomatic computing of its area. In a word, the final algorithm is convenient to track and diagnose the state of an illness for doctor in the online.Our thesis contains following parts:(1) we introduce the theory of wavelet transform and study the multi-scale wavelet transform analytical method.The multi-scale analytical method unites the accurate of fine scale and divisibility of grain scale with the image syncretic information at different scales which is suitable for the automatic or semiautomatic image division.(2) Transmutation model have been focused because of the continuous contours in image division field. Traditional snake model and degree-vector model are discussed. In further, we study the GVF model based the external energy of wavelet transform.(3) We realize the image division of gastric ulcer based the GVF model and wavelet transform with MATLAB. It is convenient to present the data straightly for computing the area of gastric ulcer.