摘要
针对传统C-V模型收敛速度慢且不完全适合灰度不均匀图像分割的问题,提出基于参数化全散度的C-V模型及其相应的快速阈值分割算法。将全散度引入传统C-V模型并获得一种改进的区域活动轮廓模型,然后,采用水平集和变分法相结合得到该模型所对应的偏微分方程,并通过数值求解该方程获得适合图像分割的快速迭代算法。实验结果表明,该方法分割效果及收敛速度明显提高,且具有较高的鲁棒性和抗噪性。
Aiming at the problems of traditional C-V model that it has slow convergence speed and is not fully suitable for segmenting the images with inhomogeneous intensity,in the paper we propose the parametric total Bregman divergence-based C-V model and its corresponding fast threshold segmentation algorithm. We introduce total Bregman divergence to traditional C-V model and obtain an improved regional activity contour model; then we adopt the combination of level set and variational method to get the partial differential equation corresponding to the model. Furthermore,by numerically solving this partial differential equation we attain fast iteration algorithm suitable for image segmentation. Experimental results show that the segmentation quality and convergence speed of the proposed algorithm have obviously improved,meanwhile it has higher robustness and anti-noise ability.
出处
《计算机应用与软件》
CSCD
2015年第12期179-183,共5页
Computer Applications and Software
基金
国家自然科学基金重点资助项目(90607008)
陕西省自然科学基金资助项目(2014JM8331
2014JQ5183
2014JM8307)
陕西省教育厅自然科学基金资助项目(2013JK1129)
西安邮电大学2013年研究生创新基金项目(ZL2013-23)
关键词
图像分割
阈值分割
C-V模型
全散度
Image segmentation
Threshold segmentation
C-V model
Total bregman divergence
