基于梯度加权的局部区域型水平集分割算法

A Local Region Level Set Method Based On Gradient Weighting For Image Segmentation

灰度不均匀现象普遍存在于自然图像和医学图像中,为了更好地解决图像中灰度不均匀现象带来的问题,通过在局部区域中使用梯度加权提出了一种新型水平集图像分割算法。图像的梯度信息具有不变性,在局部区域中利用梯度加权使本文模型对初始化轮廓具有很好的鲁棒性。同时使用梯度加权也提升了图像的对比度,改善了分割的效果。文中最后会通过一些合成图像和真实图像的分割实验来验证本文算法的有效性和鲁棒性。

Intensity inhomogeneity often exists in nature and medical images, in this paper, we proposed a novel image segmentation method with employing gradient weighting in local region to handle intensity inhomogeneous images. In general, the gradient of image is stable. The use of gradient weighting in local region will make our model more robust to initialized contours. At the same time, the employing of gradient weighting will enhance the contrast of image intensity between object and background, which will improve the performance of our method on image segmentation. Finally, some experiments on synthetic images and nature images will be shown to demonstrate the efficiency and robustness of our method.