Study On Applications Of Level Set Method In Image Segmentation

Recently, level set method combined with partial differential equation has been widely used in the image segmentation. The application of active contour model, which is one of the typical models of this method, has aroused the attention of the researchers. Active contour model uses the method of minimization of the energy function to drive the evolution curve to reach the target boundary. Because of the improvement theory for the level set and partial differential equation, this kind of method has increasingly become a more and more general tool in image segmentation field.The research work in the thesis includes two parts as follows:Active contour model is divided into two main categories, region-based active contour model control and Edge-based active contour model. The region-based active contour model controls the curves evolution through regional information, which will get the segmentation results based on the whole image, edge-based active contour model mainly drives the curve to the target boundary by gradient information. Combined the advantages of two types of models, the difference between the mean gray value of the background area and the target area is added into the energy function, then apply it to image segmentation. Compared with the general active contour model, the experiment results show that active contour model with the regional difference information has ideal effect of segmentation, faster evolution speed and higher efficiency. We can get more satisfactory segmentation results.In addition, in order to optimize the performance of the geometric active contour model, the paper proposes a new speed stop function. The improvements for speed stop function include two aspects. On the one hand, in order to be able to adaptively process images with different concentrations of noise, the paper proposes pre estimation of image noise to change speed stop function adaptively. On the other hand, the traditional speed stop function processes the image based on the image gradient information, but we will often encounter uneven gray image in practical application. Therefore, the paper added the region information to speed stop function, optimizing the performance of the traditional speed stop function in image processing.Through theoretical and experimental verification, the new speed stop function has the advantages and feasibility.