基于全局-局部策略的EMD概率主动轮廓模型

摘要

Active contour model is one of the most useful tools to solve image segmentation task. It has been widely used in recent years. In this paper, some prior works are discussed and Earth Mover's Distance(EMD) is introduced into statistical active contour model. Then a new statistical active contour model based on EMD is proposed. However, the proposed global strategy usually leads to local segmentation error. Therefore, a new global-local strategy which combines global and local scheme into EMD is raised. The proposed strategy not only extends the application of the proposed EMD model but also further improves the performance.%主动轮廓模型作为图像分割的有力工具,近年来得到了广泛的应用.在讨论前人工作的基础上,首先将基于区域直方图的概率估计方法和EMD距离引入主动轮廓模型,建立了基于EMD距离的能量泛函.为了解决图像局部的置乱现象容易产生分割误差的问题,提出一种全局-局部模型,提高分割精度,并应用于EMD距离,不仅扩大了模型的应用范围,而且使分割效果得到了进一步的提升.