GVF Snake Model And Its Application In Image Segmentation

Image segmentation has been a hot topic in the field of image processing and computer vision. Contour extraction based on active contour model as a relatively new and effective method for image segmentation. It compared with the traditional contour extraction algorithm, showing good segmentation feature. The gradient vector flow(GVF Snake) model is used as a parameter active contour model, it can provides a good solution for most computer vision problem.This study first introduces the research background and significance of image segmentation, and then is used in image segmentation of active contour model on the research present situation. We mainly introduced the basic principle of Snake model and numerical solution, and lead to the GVF Snake model, which is based on improve the external force of the Snake model. Then, the thorough analysis and research the GVF Snake model of theoretical knowledge. Finally, the improved model was applied to the actual medical skin mirror image segmentation. In this paper, the research content mainly includes the following three aspects:1. GVF Snake model problem for the calculation of the gradient vector field, first describes several methods to calculate the force field of the existing gradient. Proposed based on BFGS algorithm to compute the gradient force field, this is mainly based on the idea of quasi-Newton method to solve iterative equation. BFGS algorithm with two vector field has been given a detailed process of solving by computer simulation and numerical solution, and finally improved GVF Snake model for image processing. Theoretical analysis and experimental results show that this method can effectively solve the gradient field, and can better segment the target contour, with better structural characteristics of the object to keep the contour.2. GVF Snake model for problem set initial contour line, first introduced several common contour setting method. Proposed a combination of contour lines based on the theory of boundary tracking method automatically initialized. The first method uses boundary tracking algorithm rough segmentation, edge position to obtain valid information point, generates an initial contour after sampling. Then the improved GVF Snake model is used in image processing, experimental results show that this method can effectively GVF Snake model to set the initial outline, and can accurately obtain the target contour.3. Based on the previous two parts of GVF Snake model of theory analysis and improvement, will the improved GVF Snake model is applied to the segmentation of medical skin mirror image. First introduces the background and significance of medical image segmentation, and analyzes the characteristics of the medical image itself. Finally, the improved GVF Snake model for medical image segmentation. The experimental results show that the method can effectively extract skin mirror image contour, has good practical value.