摘要
针对彩色图像颜色空间特性,提出一种改进的边缘生长图像分割方法。首先根据平均颜色矩确定图像量化级数,色彩量化后采用边缘检测提取边缘像素集,并将这些高细节点形成边缘线,围成一个封闭的区域,最后根据颜色空间的区域距离,将初始分割区域进行合并。该方法很好的解决了使用边缘检测或区域生长所产生的不连续性和过分割问题。实验结果显示对彩色图像分割具有较好的效果。
An improved image segmentation method based on edge growing is proposed in view of the features of color space. Firstly, the method is applied to quantize the color image according to the average color moment. After that, we take edge detection to obtain the set of edge pixels, and use these high detail points to form the edge lines. And then, encircle a dosed region to finish the initial image segmentation, According to the distance of color space, we carry on the region consolidation at last. This method greatly solves ex-segmentation problem and discontinuous edge when edge detection or region growing technique is adopted. Experimented results show that it has good effect of color image segmentation.
出处
《微电子学与计算机》
CSCD
北大核心
2007年第6期194-197,共4页
Microelectronics & Computer
基金
四川省教育厅重点科研基金项目(2003A109)
关键词
颜色空间
颜色矩
边缘生长
彩色图像分割
color space
color moment
edge growing
color image segmentation
