Computer Engineering and Applications ›› 2017, Vol. 53 ›› Issue (2): 177-180.DOI: 10.3778/j.issn.1002-8331.1503-0310
Previous Articles Next Articles
YU Min, YU Fengqin, CHEN Ying
Online:
Published:
於 敏,于凤芹,陈 莹
Abstract: Image segmentation based on super-pixel similarity and statistical
Key words: super-pixel, merging criterion, color similarity, statistical characteristics
摘要: 为了解决超像素图像分割的过分割问题,提出了一种基于超像素区域颜色直方图相似性和统计特性的合并判断准则,用来合并超像素图像分割的区域。该合并准则将超像素分割结果作为区域合并的基本单元,利用基本单元的颜色相似性、空间距离大小和统计特性进行区域合并,以解决超像素图像分割中存在的过分割问题。仿真实验结果表明,该方法能有效地改善超像素过分割问题,并且用于评价分割算法的概率Rand指数和信息变化指数都有所提高。
关键词: 超像素, 合并准则, 颜色相似性, 统计特性
YU Min, YU Fengqin, CHEN Ying. Image segmentation based on super-pixel similarity and statistical[J]. Computer Engineering and Applications, 2017, 53(2): 177-180.
於 敏,于凤芹,陈 莹. 基于相似性和统计性的超像素的图像分割[J]. 计算机工程与应用, 2017, 53(2): 177-180.
0 / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: http://cea.ceaj.org/EN/10.3778/j.issn.1002-8331.1503-0310
http://cea.ceaj.org/EN/Y2017/V53/I2/177