The Research Of Color Image Edge-detection Method In The Surgery Simulation System

Edge detection is an indispensable step in the image preprocessing procedure. To the surgery simulation system, the edge-detection of the human body section image is a crucial process too, and it is the foundation of three-dimention modeling. However, both the edges and noises exist in the high frequency, in addition, the geometry edges and texture are usually presented as edges, which make the edge detection more difficult.This thesis discusses the color edge-detection methods based on the available edge detection methods. At the same time, it analyses and modifies the shortages of the GVF Snake model. And based on the advantages of the wavelet and GVF Snake model, a new color edge-detection deformable model is put forward.The main achievements of this thesis are as following:1. The thesis introduces the development of edge-detection methods, analyzes and realizes the traditional edge detection methods that based on the differential operators and wavelet transformation. At the same time, it analyzes and compares the results.2. GVF Snake model is an efficient method for contour detection of objects. In this thesis, a detailed analysis of GVF Snake model is given with emphasis on its limitations nd shortcomings. The experiments prove that the time of force field iterations is in proportion to the iterations, and the time of convergence is in proportion to convergence iterations too. The concept of "effective force field" is introduced, based on which, the orientation principle of initial contour is presented.3. On the basis of analyzing the shortages of GVF snake model, a new method based on wavelet transformation is proposed according to the orientation principle of initial contour. This method successfully resolves the problem of contour initialization, decreases computational cost and effectively reduces time complex.4. A new color images edge-detection method is put forward on the basis of the wavelet and GVF Snake model. The new method not only successfully extractes the objective contour but reserves all the merits of GVF Snake model.