基于RGB分量统计的可变区域彩色图像分割算法

Variable-domain approach for color image segmentation based on RGB statistical model

为了能够对彩色图像进行高效的分割,提出了一种可变区域的图像分割算法,利用基于图像全局RGB分量统计信息的活动轮廓模型进行曲线演化,并使用水平集表示轮廓。通过改变和缩小分割区域的策略,将分割过程分为多个阶段进行。在灰度图像的分割算法的基础上,将可变区域策略拓展到彩色图像。实验结果表明,图像中多连通区域的物体能够被准确且快速地分割出来。与现有模型相比,可以自动地完成工作而无须人工干预,并且算法快速方面有明显的改进。

In order to segment color images efficiently,this paper proposed a variable-domain algorithm based on active contour models and global statistical RGB components.Curve evolution was employed for numerical implementation with level sets method.The essential idea was to re-define the computing domain in images repeatedly,by separating the segmentation procedure into several individual phases.The variable domain strategy was extended to color images based on gray-scale algorithm.According to the experimental results,the objects with multiple connection regions can be accurately and rapidly segmented.Compared to current methods,the work can be done automatically without manual intervention and the rapidity is improved evidently.