针对图像分割中经典GAC模型无法准确分割深度凹陷、灰度不均匀的目标和容易穿越弱边界的问题, 提出了改进的GAC模型。利用图像Renyi熵、图像灰色关联度、图像的局部方差信息构造新能量函数, 扩大经典GAC模型的边界检测能力。仿真结果表明,改进模型在减少分割时间的同时能够成功分割出目标深度凹陷部分, 对弱边界、目标灰度不均匀或背景灰度不均匀也有较好的收敛效果, 其分割效果不仅优于CV模型, 也优于CV模型的两个最新改进模型(LBF和GACV)。

The problem of mismatching concave contours, weak edge and intensity inhomogeneity in traditional active contour model is studied. By combining the method of Renyi entropy segment, grey relational analysis of image, image variance with traditional GAC model, an improved model is proposed. In the way, the evolution curves can stop at boundaries accurately. Simulation results show that the proposed model can obtain better results with respect to images that have the intensity inhomogeneity in objects or backgrounds, weak edges, and/or concave contours, while significantly reducing segmentation time. Besides, it has many advantages over other two improved CV models (LBF and GACV).