An Efficient C-V Model with Wavelets

Abstract

We propose an efficient C-V model with wavelet transform to speed up the C-V model evolution. Firstly, the image is decomposed into the downsizing and approximation sub-image with wavelet transform. Then the sub-image composed with the approximation wavelet coefficient matrix is pre-segmented using C-V model to get an approximation curve in a short time. Thirdly, in order to get a fine segmentation, the original image is segmented subtly using C-V model again, of which the initial curve is reconstructed by the result of pre-segmentation. The experimental results validate the effectiveness and feasibility of this model.