摘要
本文采用分水岭算法对红外图像进行分割 ,针对其存在的过分割问题以及红外图像的特点 ,提出了分割区域边界平均灰度及其面积对过分割区域进行合并 ,以得到有意义的分割结果。本算法利用空中红外目标图像进行了实验。实验表明 ,本算法适用于红外目标图像 。
In this paper,an algorithm based on watershed transformation has been applied,which is adapted to infrared image segmentation.According to the over segmentation of watershed algorithm and the characters of infrared image,the gray means of the region and the region areas,have been presented,which were used t unify these over segmented regions into some significant object regions.Finally,several experiments to verify the proposed method have been given with good results.
出处
《光电子．激光》
EI
CAS
CSCD
北大核心
2001年第10期1072-1075,共4页
Journal of Optoelectronics·Laser
基金
2 0 7所国防重点实验室资助项目 (0 0 JS6 6 .3.1.BQ.0 110 )
关键词
图像分割
分水岭算法
红外图像
image segmenttation
watershed algorithm
over segmentation
