In image segmentation methods,we often take a clear image as the object.We can get high quality segmentation in experiments.In real application images always have noise.Many existed segmentation methods for high noise image are not satisfactory.A segmentation method coupled with denoising is proposed in this thesis.In the coupled method,a high noise image is first denoised by using a new variational model,then the denoised image is segmented by the improved CV model.In the numerical methods,the multiple directions differences are used to approximate the partial derivatives.The numerical experimental results show that the proposed coupling denoising and segmentation method can show its validity,especially for high noise of the image segmentation has obvious superiority compared with CV model. |