局部信息熵的快速混合测地区域活动轮廓模型

Fast Hybrid Geodesic Region Active Contour Model of Local Entropy

针对变分水平集算法在图像分割过程中计算量较大且收敛速度慢的现象,在前人研究的基础上提出了一种新的局部信息熵的混合测地区域活动轮廓模型。该模型构造一个新的能量泛函,在泛函中引入柔化核函数作为窗口核函数,构造一个新的符号压力函数来代替测地线边缘检测函数,并以局部信息熵作为图像拟合能量项的权重,通过非凸正则化项来约束水平集函数。由此得到的算法不仅能加快轮廓曲线的收敛速度,而且可以处理那些由于光照或其他外界因素的变化产生的灰度不均匀或者模糊的图像,提高分割的精确性。将算法在合成图像和真实图像上做仿真实验,实验结果表明,该算法具有较快的收敛速度,分割也较准确,同时对轮廓曲线的初始位置不敏感,具有很好的鲁棒性。

Aiming at the phenomenon that the variational level set algorithm takes large amount of calculation and converges slowly in the process of image segmentation, this paper proposes a new hybrid geodesic region active contour model of local entropy based on the predecessors' research. The model constructs a new energy functional, introduces a mollifying kernel function to be window function, constructs a new signed pressure force function to replace the geodesic edge stopping function, and uses local entropy as the weight of image fitting energy, then adds a nonconvex regularization term to constrain the level set function. The algorithm getting from this not only accelerates the convergence rate of the contour curve, but also can address image segmentation inaccuracy for image intensity inhomogeneity or blurring caused by the change of illumination or other external factors. The simulation experiment results on synthetic images and real images show that the proposed algorithm is of higher converging speed and better accuracy, less sensitive to the location of initial contour at the same time, and has better robustness.